1_MasteringSpacy_ch3_10.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "7b309e09",
   "metadata": {},
   "source": [
    "# Chapter 3: Linguistic features"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dd06ec6d",
   "metadata": {},
   "source": [
    "### Parts of Speech Tags"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6f8edd03",
   "metadata": {},
   "outputs": [],
   "source": [
    "import spacy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "8b606d6e-141b-46e9-8007-e7f3ed041f78",
   "metadata": {},
   "outputs": [],
   "source": [
    "nlp = spacy.load(\"en_core_web_sm\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "9ce513b2-deef-485b-aff1-0e71d2b39c6c",
   "metadata": {},
   "outputs": [],
   "source": [
    "doc = nlp(\"I will ship the package tomorrow.\") "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "5af55626-0413-4f2b-870f-ba87faf5f4df",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "I   tag: PRP   explain: pronoun, personal\n",
      "will   tag: MD   explain: verb, modal auxiliary\n",
      "ship   tag: VB   explain: verb, base form\n",
      "the   tag: DT   explain: determiner\n",
      "package   tag: NN   explain: noun, singular or mass\n",
      "tomorrow   tag: NN   explain: noun, singular or mass\n",
      ".   tag: .   explain: punctuation mark, sentence closer\n"
     ]
    }
   ],
   "source": [
    "#spacy.explain provides parts of speech tag explanations\n",
    "\n",
    "for token in doc:\n",
    "    print(f\"{token.text}  tag: {token.tag_}  explain: {spacy.explain(token.tag_)}\") "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "67915733",
   "metadata": {},
   "source": [
    "### Visualize dependency labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "afab2676",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<span class=\"tex2jax_ignore\"><svg xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\" xml:lang=\"en\" id=\"0d0fc20dfe5843fd876f256de0f06a35-0\" class=\"displacy\" width=\"925\" height=\"399.5\" direction=\"ltr\" style=\"max-width: none; height: 399.5px; color: #000000; background: #ffffff; font-family: Arial; direction: ltr\">\n",
       "<text class=\"displacy-token\" fill=\"currentColor\" text-anchor=\"middle\" y=\"309.5\">\n",
       "    <tspan class=\"displacy-word\" fill=\"currentColor\" x=\"50\">I</tspan>\n",
       "    <tspan class=\"displacy-tag\" dy=\"2em\" fill=\"currentColor\" x=\"50\">PRON</tspan>\n",
       "</text>\n",
       "\n",
       "<text class=\"displacy-token\" fill=\"currentColor\" text-anchor=\"middle\" y=\"309.5\">\n",
       "    <tspan class=\"displacy-word\" fill=\"currentColor\" x=\"225\">own</tspan>\n",
       "    <tspan class=\"displacy-tag\" dy=\"2em\" fill=\"currentColor\" x=\"225\">VERB</tspan>\n",
       "</text>\n",
       "\n",
       "<text class=\"displacy-token\" fill=\"currentColor\" text-anchor=\"middle\" y=\"309.5\">\n",
       "    <tspan class=\"displacy-word\" fill=\"currentColor\" x=\"400\">a</tspan>\n",
       "    <tspan class=\"displacy-tag\" dy=\"2em\" fill=\"currentColor\" x=\"400\">DET</tspan>\n",
       "</text>\n",
       "\n",
       "<text class=\"displacy-token\" fill=\"currentColor\" text-anchor=\"middle\" y=\"309.5\">\n",
       "    <tspan class=\"displacy-word\" fill=\"currentColor\" x=\"575\">ginger</tspan>\n",
       "    <tspan class=\"displacy-tag\" dy=\"2em\" fill=\"currentColor\" x=\"575\">NOUN</tspan>\n",
       "</text>\n",
       "\n",
       "<text class=\"displacy-token\" fill=\"currentColor\" text-anchor=\"middle\" y=\"309.5\">\n",
       "    <tspan class=\"displacy-word\" fill=\"currentColor\" x=\"750\">cat.</tspan>\n",
       "    <tspan class=\"displacy-tag\" dy=\"2em\" fill=\"currentColor\" x=\"750\">NOUN</tspan>\n",
       "</text>\n",
       "\n",
       "<g class=\"displacy-arrow\">\n",
       "    <path class=\"displacy-arc\" id=\"arrow-0d0fc20dfe5843fd876f256de0f06a35-0-0\" stroke-width=\"2px\" d=\"M70,264.5 C70,177.0 215.0,177.0 215.0,264.5\" fill=\"none\" stroke=\"currentColor\"/>\n",
       "    <text dy=\"1.25em\" style=\"font-size: 0.8em; letter-spacing: 1px\">\n",
       "        <textPath xlink:href=\"#arrow-0d0fc20dfe5843fd876f256de0f06a35-0-0\" class=\"displacy-label\" startOffset=\"50%\" side=\"left\" fill=\"currentColor\" text-anchor=\"middle\">nsubj</textPath>\n",
       "    </text>\n",
       "    <path class=\"displacy-arrowhead\" d=\"M70,266.5 L62,254.5 78,254.5\" fill=\"currentColor\"/>\n",
       "</g>\n",
       "\n",
       "<g class=\"displacy-arrow\">\n",
       "    <path class=\"displacy-arc\" id=\"arrow-0d0fc20dfe5843fd876f256de0f06a35-0-1\" stroke-width=\"2px\" d=\"M420,264.5 C420,89.5 745.0,89.5 745.0,264.5\" fill=\"none\" stroke=\"currentColor\"/>\n",
       "    <text dy=\"1.25em\" style=\"font-size: 0.8em; letter-spacing: 1px\">\n",
       "        <textPath xlink:href=\"#arrow-0d0fc20dfe5843fd876f256de0f06a35-0-1\" class=\"displacy-label\" startOffset=\"50%\" side=\"left\" fill=\"currentColor\" text-anchor=\"middle\">det</textPath>\n",
       "    </text>\n",
       "    <path class=\"displacy-arrowhead\" d=\"M420,266.5 L412,254.5 428,254.5\" fill=\"currentColor\"/>\n",
       "</g>\n",
       "\n",
       "<g class=\"displacy-arrow\">\n",
       "    <path class=\"displacy-arc\" id=\"arrow-0d0fc20dfe5843fd876f256de0f06a35-0-2\" stroke-width=\"2px\" d=\"M595,264.5 C595,177.0 740.0,177.0 740.0,264.5\" fill=\"none\" stroke=\"currentColor\"/>\n",
       "    <text dy=\"1.25em\" style=\"font-size: 0.8em; letter-spacing: 1px\">\n",
       "        <textPath xlink:href=\"#arrow-0d0fc20dfe5843fd876f256de0f06a35-0-2\" class=\"displacy-label\" startOffset=\"50%\" side=\"left\" fill=\"currentColor\" text-anchor=\"middle\">compound</textPath>\n",
       "    </text>\n",
       "    <path class=\"displacy-arrowhead\" d=\"M595,266.5 L587,254.5 603,254.5\" fill=\"currentColor\"/>\n",
       "</g>\n",
       "\n",
       "<g class=\"displacy-arrow\">\n",
       "    <path class=\"displacy-arc\" id=\"arrow-0d0fc20dfe5843fd876f256de0f06a35-0-3\" stroke-width=\"2px\" d=\"M245,264.5 C245,2.0 750.0,2.0 750.0,264.5\" fill=\"none\" stroke=\"currentColor\"/>\n",
       "    <text dy=\"1.25em\" style=\"font-size: 0.8em; letter-spacing: 1px\">\n",
       "        <textPath xlink:href=\"#arrow-0d0fc20dfe5843fd876f256de0f06a35-0-3\" class=\"displacy-label\" startOffset=\"50%\" side=\"left\" fill=\"currentColor\" text-anchor=\"middle\">dobj</textPath>\n",
       "    </text>\n",
       "    <path class=\"displacy-arrowhead\" d=\"M750.0,266.5 L758.0,254.5 742.0,254.5\" fill=\"currentColor\"/>\n",
       "</g>\n",
       "</svg></span>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import spacy\n",
    "from spacy import displacy\n",
    "\n",
    "nlp = spacy.load(\"en_core_web_md\")\n",
    "\n",
    "doc = nlp(\"I own a ginger cat.\")\n",
    "\n",
    "displacy.render(doc, style= \"dep\") #dep style parameter\n",
    "\n",
    "#filename = \"cat_line.svg\" \n",
    "#output_path = Path (\"/Learning/spaCy\" + file_name)\n",
    "#output_path.open(\"w\", encoding = \"utf-8\").write(svg)\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fa32a1df",
   "metadata": {},
   "source": [
    "### Extracting named entities from website"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "7e5b9d10-9781-4ada-b36e-b73c54a80e30",
   "metadata": {},
   "outputs": [],
   "source": [
    "from bs4 import BeautifulSoup \n",
    "import requests \n",
    "import spacy \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "6e682ac5-855a-4ec9-bb98-e1f0a6cf07f0",
   "metadata": {},
   "outputs": [],
   "source": [
    "def url_text(url_string):\n",
    "    res = requests.get(url_string)\n",
    "    html = res.text\n",
    "    soup = BeautifulSoup(html, 'html5lib')\n",
    "    for script in soup([\"script\", \"style\", 'aside']):\n",
    "        script.extract()\n",
    "        text = soup.get_text()\n",
    "        return \" \".join(text.split()) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "c7db751d-228c-43fd-a777-ae04370b976a",
   "metadata": {},
   "outputs": [],
   "source": [
    "herbs = url_text(\"https://www.henriettes-herb.com/eclectic/kings/abies-cana.html\") \n",
    "nlp = spacy.load(\"en_core_web_md\") \n",
    "doc = nlp(herbs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "4b4e9733-935e-4f70-819b-e0a5cf841264",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "64"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(doc.ents)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "a669bced-acb1-4d87-8e78-5c05eed67632",
   "metadata": {},
   "outputs": [],
   "source": [
    "from collections import Counter"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "96171a02-d819-4361-8fa4-6fbeb341a9e6",
   "metadata": {},
   "outputs": [],
   "source": [
    "#counting entity labels\n",
    "labels = [ent.label_ for ent in doc.ents]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "99df83dc-b681-488a-b6cf-6eb7e1cff3a8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Counter({'ORG': 15,\n",
       "         'PERSON': 14,\n",
       "         'DATE': 7,\n",
       "         'GPE': 7,\n",
       "         'NORP': 1,\n",
       "         'QUANTITY': 3,\n",
       "         'CARDINAL': 14,\n",
       "         'ORDINAL': 1,\n",
       "         'TIME': 1,\n",
       "         'WORK_OF_ART': 1})"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Counter(labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "20d304f7-16ab-4084-aa20-c43883ca539a",
   "metadata": {},
   "outputs": [],
   "source": [
    "#counting the tokens\n",
    "items = [ent.text for ent in doc.ents]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "d9effb2d-dea3-47f8-b98c-6a672f484248",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('5', 3),\n",
       " ('American Dispensatory', 2),\n",
       " ('1898', 2),\n",
       " ('Tsuga', 2),\n",
       " ('Canada', 2),\n",
       " ('Kennedy', 2),\n",
       " ('10', 2),\n",
       " ('2', 2),\n",
       " ('Henriette Kress', 2),\n",
       " ('Abies Canadensis.—Hemlock Spruce', 1),\n",
       " (\"Henriette's\", 1),\n",
       " (\"Luentoja King's\", 1),\n",
       " ('Abies Nigra.—Black Spruce - Pix Burgundica', 1),\n",
       " ('Abies', 1),\n",
       " ('Michaux', 1),\n",
       " ('Linné', 1),\n",
       " ('Picea', 1),\n",
       " ('Hemlock', 1),\n",
       " ('above 75 feet', 1),\n",
       " ('about half an inch', 1),\n",
       " ('two', 1),\n",
       " ('Chemical Composition.—The Hemlock', 1),\n",
       " ('the Northern States', 1),\n",
       " ('second', 1),\n",
       " ('Phil', 1)]"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Counter(items).most_common(25)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "7a748751-5616-4ed9-b27a-5d64076e473d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(Abies Canadensis.—Hemlock Spruce, Henriette's, Luentoja King's, American Dispensatory, 1898, Tsuga, Abies Nigra.—Black Spruce - Pix Burgundica, Abies, Michaux, Linné, Picea, Tsuga, Hemlock, above 75 feet, about half an inch, two, Chemical Composition.—The Hemlock, the Northern States, Canada, Canada, second, Phil, Vol, II, 20, Schimmel & Co., Semi-Annual, Oct., 1893, 0.907, 15°, 59° F., 54, Two, Kennedy, Kennedy, Medical Uses, ani, W. K. Everson, morbus, 5, 10, 10 or 20 minutes, 5, 30, 2, 10, 2, 5, American Dispensatory, 1898, Harvey Wickes Felter, M.D., John Uri Lloyd, Phr, Abies Nigra.—Black Spruce, today, Distanskurs Lär, Classic Texts Archives, Henriette, 1995, Henriette Kress, Helsinki, Finland, 1995–2021, Henriette Kress)\n"
     ]
    }
   ],
   "source": [
    "print(doc.ents)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3c9d6af4",
   "metadata": {},
   "source": [
    "### displaCy Named Entity Visualizer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "efc9ec4e",
   "metadata": {},
   "outputs": [],
   "source": [
    "import spacy\n",
    "from spacy import displacy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "3c879716",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<span class=\"tex2jax_ignore\"><div class=\"entities\" style=\"line-height: 2.5; direction: ltr\">This study pointed to the assessment of the chemical composition (F, Ca, Mg, Na, K, Fe, \n",
       "<mark class=\"entity\" style=\"background: #feca74; padding: 0.45em 0.6em; margin: 0 0.25em; line-height: 1; border-radius: 0.35em;\">\n",
       "    Mn\n",
       "    <span style=\"font-size: 0.8em; font-weight: bold; line-height: 1; border-radius: 0.35em; vertical-align: middle; margin-left: 0.5rem\">GPE</span>\n",
       "</mark>\n",
       ", \n",
       "<mark class=\"entity\" style=\"background: #feca74; padding: 0.45em 0.6em; margin: 0 0.25em; line-height: 1; border-radius: 0.35em;\">\n",
       "    Zn\n",
       "    <span style=\"font-size: 0.8em; font-weight: bold; line-height: 1; border-radius: 0.35em; vertical-align: middle; margin-left: 0.5rem\">GPE</span>\n",
       "</mark>\n",
       ", \n",
       "<mark class=\"entity\" style=\"background: #feca74; padding: 0.45em 0.6em; margin: 0 0.25em; line-height: 1; border-radius: 0.35em;\">\n",
       "    Cu\n",
       "    <span style=\"font-size: 0.8em; font-weight: bold; line-height: 1; border-radius: 0.35em; vertical-align: middle; margin-left: 0.5rem\">GPE</span>\n",
       "</mark>\n",
       ", Ni, Co, Cr, Cd, and carbohydrate) of different marine seaweeds (red, green, and brown) from the \n",
       "<mark class=\"entity\" style=\"background: #c887fb; padding: 0.45em 0.6em; margin: 0 0.25em; line-height: 1; border-radius: 0.35em;\">\n",
       "    Egyptian\n",
       "    <span style=\"font-size: 0.8em; font-weight: bold; line-height: 1; border-radius: 0.35em; vertical-align: middle; margin-left: 0.5rem\">NORP</span>\n",
       "</mark>\n",
       " \n",
       "<mark class=\"entity\" style=\"background: #ff9561; padding: 0.45em 0.6em; margin: 0 0.25em; line-height: 1; border-radius: 0.35em;\">\n",
       "    Mediterranean Sea\n",
       "    <span style=\"font-size: 0.8em; font-weight: bold; line-height: 1; border-radius: 0.35em; vertical-align: middle; margin-left: 0.5rem\">LOC</span>\n",
       "</mark>\n",
       " coast. The results showed that green seaweeds supplied better calcium sources than the red and brown ones.Also, red and brown seaweeds showed higher averages of Na and K than that in green species and these seaweeds could play an important role in the electrolyte balance in humans.</div></span>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "nlp = spacy.load(\"en_core_web_sm\")\n",
    "\n",
    "text = \"This study pointed to the assessment of the chemical composition (F, Ca, Mg, Na, K, Fe, Mn, Zn, Cu, Ni, Co, Cr, Cd, and carbohydrate) of different marine seaweeds (red, green, and brown) from the Egyptian Mediterranean Sea coast. The results showed that green seaweeds supplied better calcium sources than the red and brown ones.Also, red and brown seaweeds showed higher averages of Na and K than that in green species and these seaweeds could play an important role in the electrolyte balance in humans.\"\n",
    "doc = nlp(text)\n",
    "displacy.render(doc, style=\"ent\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "65857703",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mn 384\n",
      "Zn 384\n",
      "Cu 384\n",
      "Egyptian 381\n",
      "Mediterranean Sea 385\n"
     ]
    }
   ],
   "source": [
    "## Predicting named entities in context\n",
    "\n",
    "import spacy\n",
    "nlp = spacy.load(\"en_core_web_sm\")\n",
    "\n",
    "text = \"This study pointed to the assessment of the chemical composition (F, Ca, Mg, Na, K, Fe, Mn, Zn, Cu, Ni, Co, Cr, Cd, and carbohydrate) of different marine seaweeds (red, green, and brown) from the Egyptian Mediterranean Sea coast. The results showed that green seaweeds supplied better calcium sources than the red and brown ones.Also, red and brown seaweeds showed higher averages of Na and K than that in green species and these seaweeds could play an important role in the electrolyte balance in humans.\"\n",
    "doc = nlp(text)\n",
    "\n",
    "for ent in doc.ents:\n",
    "    print(ent.text, ent.label)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2057230d-ab7e-4bcd-8ada-36db7baa4d4d",
   "metadata": {},
   "source": [
    "### Grouping a multi-word token"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0befa920-0732-4b19-ac53-061f015dc45e",
   "metadata": {},
   "outputs": [],
   "source": [
    "doc = nlp(\"She lived in New Hampshire\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "644efbba-dec4-41f1-8a2c-e514b9e88890",
   "metadata": {},
   "outputs": [],
   "source": [
    "doc.ents"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3c8c2644-c7b3-4113-8219-1b306c4b0a9d",
   "metadata": {},
   "outputs": [],
   "source": [
    "[(token.text, token.i) for token in doc] #New and Hampshire not captured as one token"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "139a2176-27a8-4e82-b7f8-8669493c66a2",
   "metadata": {},
   "outputs": [],
   "source": [
    "len(doc)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "78251389-6da7-4c29-8ad7-6c68ed241857",
   "metadata": {},
   "outputs": [],
   "source": [
    "#doc.retokenize() correct tool for merging and splitting spans. Here merging:\n",
    "with doc.retokenize() as retokenizer:\n",
    "    retokenizer.merge(doc[3:5])\n",
    "    attrs = {\"LEMMA\": \"new hampshire\"}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "af46eb02-b53e-4cd3-9642-07594a5862ce",
   "metadata": {},
   "outputs": [],
   "source": [
    "[(token.text, token.i) for token in doc]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "588d5308-1bdb-4ce7-aa3f-5e5a746fcaac",
   "metadata": {},
   "source": [
    "### Breaking apart a multi-word token"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "a2cea173-446d-41b0-b73e-786e891279f7",
   "metadata": {},
   "outputs": [],
   "source": [
    "doc = nlp(\"She live in NewHampshire\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5ec38425-4b66-4516-95ab-9c37e2327fa3",
   "metadata": {},
   "outputs": [],
   "source": [
    "len(doc)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "968de5b8-fa48-4bc4-bf53-2a30a00d9d46",
   "metadata": {},
   "outputs": [],
   "source": [
    "for token in doc:\n",
    "    print(token.text, token.pos_, token.tag_, token.dep_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "378e0fcb-5245-4a97-8e02-6953df7ccc29",
   "metadata": {},
   "outputs": [],
   "source": [
    "#doc.retokenize() correct tool for merging and splitting spans.\n",
    "with doc.retokenize() as retokenizer:\n",
    "    heads = [(doc[3], 1), doc[2]] #doc[3] will be the +1th position token in new setup\n",
    "    attrs = {\"TAG\": [\"NNP\", \"NNP\"], \"DEP\": [\"pobj\", \"compound\"]}\n",
    "    retokenizer.split(doc[3], [\"New\", \"Hampshire\"], heads=heads, attrs=attrs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "410c0f9f-9f84-4c48-9060-abd384ecbd14",
   "metadata": {},
   "outputs": [],
   "source": [
    "[(token.text, token.lemma_, token.i) for token in doc]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e10a4009-edde-424a-8972-dfc25ef86d13",
   "metadata": {},
   "outputs": [],
   "source": [
    "for token in doc:\n",
    "    print(token.text, token.pos_, token.tag_, token.dep_)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cbe0ad6f",
   "metadata": {},
   "source": [
    "# Chapter 4"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b6d5481a-05b2-4136-9c86-a8d72286b306",
   "metadata": {},
   "source": [
    "### Using the Matcher"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "503db712-bbe4-4400-bb96-4435c866f193",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Matcher class can match our predefined rules to the sequesnce of tokens in Doc and Span objects"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "27cf4bc1-f13f-45e6-a454-be54c9d20d9e",
   "metadata": {},
   "outputs": [],
   "source": [
    "import spacy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "3ddb40fe-050d-4a41-a042-b678d494e947",
   "metadata": {},
   "outputs": [],
   "source": [
    "from spacy.matcher import Matcher"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fcc48e78-e564-47d6-bfb9-fe37d35afd1e",
   "metadata": {},
   "outputs": [],
   "source": [
    "nlp = spacy.load(\"en_core_web_sm\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "1d724566-9c85-46d3-a5b9-d2d481eea7cc",
   "metadata": {},
   "outputs": [],
   "source": [
    "doc = nlp(\"Good morning, I want to reserve a ticket. I will then say good evening!\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "47041e99-ff53-4b72-9845-e5ed2558cf02",
   "metadata": {},
   "outputs": [],
   "source": [
    "#matcher needs to be initialized with vocabulary object\n",
    "matcher = Matcher(nlp.vocab)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "9aece0c1-aa85-4ec9-9fc6-e34ec46085c0",
   "metadata": {},
   "outputs": [],
   "source": [
    "pattern1 = [{\"LOWER\": \"good\"}, {\"LOWER\": \"morning\"}, {\"IS_PUNCT\": True}]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "819a0a84-8d45-4331-ae39-256a4606c4da",
   "metadata": {},
   "outputs": [],
   "source": [
    "#introduce pattern to matcher\n",
    "matcher.add(\"morningGreeting\", [pattern1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "2c277963-66e5-4eff-a4a7-8799c8af9c00",
   "metadata": {},
   "outputs": [],
   "source": [
    "pattern2 = [{\"LOWER\": \"good\"},{\"LOWER\": \"evening\"}, {\"IS_PUNCT\": True}]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "9def8ad7-18fc-4136-8070-1520dc92b061",
   "metadata": {},
   "outputs": [],
   "source": [
    "matcher.add(\"eveningGreeting\", [pattern2])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "cc69f3b5-5dea-447c-b977-cd6fba75132c",
   "metadata": {},
   "outputs": [],
   "source": [
    "matches = matcher(doc)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "a6cf8610-6ab0-4b94-be59-ff786f6e6ebd",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 3 Good morning,\n",
      "14 17 good evening!\n"
     ]
    }
   ],
   "source": [
    "for match_id, start, end in matches:\n",
    "    pattern_name = nlp.vocab.strings[match_id]\n",
    "    m_span = doc[start:end]\n",
    "    print(start, end, m_span.text)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "daf65e39-97e6-4125-a953-5d7826cb4529",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Example 2 Pattern on 2 sentences + other attributes\n",
    "doc1 = nlp(\"Can you swim?\")\n",
    "doc2 = nlp(\"Can Sally swim?\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c43ee697-bb38-437b-a793-8972c4377070",
   "metadata": {},
   "outputs": [],
   "source": [
    "pattern = [{\"IS_SENT_START\": True, \"LOWER\": \"can\"}, {\"IS_TITLE\": True}] #put two attributes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e1362cd5-99ba-4239-afe9-b7f177505b0c",
   "metadata": {},
   "outputs": [],
   "source": [
    "matcher.add(\"canThenCapitalized\", [pattern])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b506ef0e-2241-4586-94e0-248d477af1a7",
   "metadata": {},
   "outputs": [],
   "source": [
    "matcher(doc)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "cf70e4e4-6630-48ee-9ba5-3b3f9a6640bb",
   "metadata": {},
   "outputs": [],
   "source": [
    "matches = matcher(doc2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0452e3cd-cc00-42f5-a103-f78a7dda9010",
   "metadata": {},
   "outputs": [],
   "source": [
    "len(matches)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1d204e65-1fdd-41df-a297-15eb3d104ff0",
   "metadata": {},
   "outputs": [],
   "source": [
    "mid, start, end = matches[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7764de87-287b-44c3-a15f-206e3f1bf202",
   "metadata": {},
   "outputs": [],
   "source": [
    "print(start, end, doc2[start:end]) \n",
    "#prints only first token in sentence is can and second word has a capitalized letter"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d2ec08db-5318-49c9-bf9b-009e85c6d57a",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Example 3 - One way of finding yes/no question sentences\n",
    "doc = nlp(\"Will you go there?\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0f2a7b6c-dd87-4a4e-b2d2-0f22eb6afc63",
   "metadata": {},
   "outputs": [],
   "source": [
    "pattern = [{\"IS_SENT_START\": True, \"TAG\": \"MD\"}]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a26f4d3f-e9cc-4a7b-9ba0-85a3f57c3a1d",
   "metadata": {},
   "outputs": [],
   "source": [
    "matcher.add(\"sentStartingAuxVerb\", [pattern])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5cbc3a37-d3b7-4ec7-b48e-0d2accb16ab7",
   "metadata": {},
   "outputs": [],
   "source": [
    "matches = matcher(doc)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "51449037-f8b4-4567-830e-c1fe6a0642f3",
   "metadata": {},
   "outputs": [],
   "source": [
    "len(matches)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3b92c8ec-85ac-4d31-b507-8082a440300d",
   "metadata": {},
   "outputs": [],
   "source": [
    "mid, start, end = matches[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "76172c9d-f49f-4ba8-942b-c059bcd3bc72",
   "metadata": {},
   "outputs": [],
   "source": [
    "print(start, end, doc[start:end])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b1ad105f-af03-4629-98a5-d545c5182c5a",
   "metadata": {},
   "outputs": [],
   "source": [
    "doc2 = nlp(\"I might go there.\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a36f38be-f4b8-40a0-93ce-9a5ff27750a9",
   "metadata": {},
   "outputs": [],
   "source": [
    "matcher(doc2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "590305f7-3e42-4c3d-8f1f-5716656a5ed8",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Example 4 - using extended syntax - using in\n",
    "doc = nlp(\"Good morning, I'm here. I'll say good evening!!\") \n",
    "\n",
    "pattern = [{\"LOWER\": \"good\"},{\"LOWER\": {\"IN\": [\"morning\", \"evening\"]}},\n",
    "           {\"IS_PUNCT\": True}] \n",
    "\n",
    "matcher.add(\"greeting\",[pattern]) \n",
    "\n",
    "matches = matcher(doc) \n",
    "\n",
    "for mid, start, end in matches:\n",
    "    print(start, end, doc[start:end]) \n",
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "22e5a8f9-aaf1-4bd7-bd85-e5b62030b334",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Example 4 - using extended syntax - regex ? (zero or one times)\n",
    "doc1 = nlp(\"Barack Obama visited France.\") \n",
    "doc2 = nlp(\"Barack Hussein Obama visited France.\") \n",
    "pattern = [{\"LOWER\": \"barack\"}, {\"LOWER\": \"hussein\", \"OP\": \"?\"},\n",
    "           {\"LOWER\": \"obama\"}] \n",
    "\n",
    "matcher.add(\"obamaNames\", [pattern]) \n",
    "matcher(doc1)  \n",
    "matcher(doc2) \n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9203dbf5-840d-4164-bbf1-45c1d4e0c1de",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Example 5, using * (zero or more times)\n",
    "doc1 = nlp(\"Hello hello hello, how are you?\") \n",
    "doc2 = nlp(\"Hello, how are you?\") \n",
    "doc3 = nlp(\"How are you?\") \n",
    "\n",
    "pattern = [{\"LOWER\": {\"IN\": [\"hello\", \"hi\", \"hallo\"]}, \"OP\":\"*\"}, {\"IS_PUNCT\": True}] \n",
    "\n",
    "matcher.add(\"greetings\", [pattern]) \n",
    "            \n",
    "for mid, start, end in matcher(doc1):\n",
    "            print(start, end, doc1[start:end]) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e0c97703-897c-416d-8fec-c316b42af8bd",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Example 6 using the wildcard\n",
    "doc = nlp(\"My name is Alice and his name was Elliot.\") \n",
    "\n",
    "pattern = [{\"LOWER\": \"name\"},{\"LEMMA\": \"be\"},{}] \n",
    "matcher.add(\"pickName\", [pattern]) \n",
    "\n",
    "for mid, start, end in matcher(doc):\n",
    "    print(start, end, doc[start:end]) \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ecb97242-ad76-4add-b122-9a7ac9d37a40",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Example 7 using the wildcard to ignore a token\n",
    "\n",
    "doc1 = nlp(\"I forwarded his email to you.\") \n",
    "doc2 = nlp(\"I forwarded an email to you.\") \n",
    "doc3 = nlp(\"I forwarded the email to you.\") \n",
    "\n",
    "pattern = [{\"LEMMA\": \"forward\"}, {}, {\"LOWER\": \"email\"}] \n",
    "matcher.add(\"forwardMail\", [pattern]) \n",
    "\n",
    "for mid, start, end in matcher(doc1):\n",
    "    print(start, end, doc1[start:end]) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "bfdf433a-5416-479a-a083-3f396b274437",
   "metadata": {},
   "outputs": [],
   "source": [
    "for mid, start, end in matcher(doc2):\n",
    "    print(start, end, doc2[start:end]) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2d30762e-ae6a-4358-9201-684f9e1da43f",
   "metadata": {},
   "outputs": [],
   "source": [
    "for mid, start, end in matcher(doc3):\n",
    "    print(start, end, doc3[start:end]) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ceb263b5-45f8-4903-8073-7bb802a5724c",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Example 8 Regex support -  selecting variations of traveled\n",
    "doc1 = nlp(\"I travelled by bus.\") \n",
    "doc2 = nlp(\"She traveled by bike.\") \n",
    "\n",
    "pattern = [{\"POS\": \"PRON\"}, {\"TEXT\": {\"REGEX\": \"[Tt]ravell?ed\"}}] \n",
    "\n",
    "matcher.add(\"travelRegex\", [pattern]) \n",
    "\n",
    "for mid, start, end in matcher(doc1): \n",
    "    print(start, end, doc1[start:end]) \n",
    "\n",
    "for mid, start, end in matcher(doc2): \n",
    "    print(start, end, doc2[start:end]) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "386616f7-9505-44b1-ae35-972c1764066a",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Example 9 - using regex to select all verbs\n",
    "doc = nlp(\"I went to Italy; he has been there too. His mother also has told me she wants to visit Rome.\") \n",
    "\n",
    "pattern = [{\"TAG\": {\"REGEX\": \"^V\"}}] \n",
    "\n",
    "matcher.add(\"verbs\", [pattern])\n",
    "\n",
    "for mid, start, end in matcher(doc):\n",
    "    print(start, end, doc[start:end]) \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2d8ff711-ad12-464c-acfe-b6eebf766d2d",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Example 10 - Using PhraseMatcher to match long dictionaries\n",
    "import spacy \n",
    "from spacy.matcher import PhraseMatcher \n",
    "\n",
    "nlp = spacy.load(\"en_core_web_sm\") \n",
    "\n",
    "matcher = PhraseMatcher(nlp.vocab) \n",
    "\n",
    "terms = [\"Angela Merkel\", \"Donald Trump\", \"Alexis Tsipras\"] \n",
    "\n",
    "patterns = [nlp.make_doc(term) for term in terms] \n",
    "\n",
    "matcher.add(\"politiciansList\", None, *patterns) \n",
    "\n",
    "doc = nlp(\"3 EU leaders met in Berlin. German chancellor Angela Merkel first welcomed the US president Donald Trump. The following day Alexis Tsipras joined them in Brandenburg.\")\n",
    "\n",
    "matches = matcher(doc) \n",
    "\n",
    "for mid, start, end in matches: \n",
    "    print(start, end, doc[start:end])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "35afff5d-ebc6-4cfe-8354-88d1bd08e5ba",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Example 11 Matching Shape attribute\n",
    "matcher = PhraseMatcher(nlp.vocab, attr=\"SHAPE\")\n",
    "\n",
    "ip_nums = [\"127.0.0.0\", \"127.256.0.0\"] \n",
    "patterns = [nlp.make_doc(ip) for ip in ip_nums]\n",
    "\n",
    "matcher.add(\"IPNums\", None, *patterns) \n",
    "\n",
    "doc = nlp(\"This log contains the following IP addresses: 192.1.1.1 and 192.12.1.1 and 192.160.1.1 .\") \n",
    "\n",
    "for mid, start, end in matcher(doc):\n",
    "    print(start, end, doc[start:end]) \n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "60659402-9277-4f0a-973f-cce6891232cd",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Example 12 Matching entity and verb/action\n",
    "\n",
    "pattern = [{\"ENT_TYPE\": \"PERSON\", \"OP\": \"+\"}, {\"POS\" : \"VERB\"}]\n",
    "\n",
    "matcher.add(\"personEntAction\", [pattern])\n",
    "\n",
    "doc = nlp(\"Today German chancellor Angela Merkel met with the US president.\") \n",
    "\n",
    "matches = matcher(doc)\n",
    "\n",
    "for mid, start, end in matches:\n",
    "    print(start, end, doc[start:end]) "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "762f3a44-53f9-478e-b9e1-4f8893f7e682",
   "metadata": {},
   "source": [
    "### Rule Based Entity Extraction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0bdb0c31-d868-4440-91aa-4b41bca113af",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Entity Ruler allows us to add rules on top of the stat model and creates a more powerful NER model.\n",
    "#ER is a pipeline\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "09522c4c-fc5d-4b94-97f4-a39f8448b909",
   "metadata": {},
   "outputs": [],
   "source": [
    "import spacy\n",
    "from spacy.pipeline import EntityRuler\n",
    "\n",
    "nlp = spacy.load(\"en_core_web_md\")\n",
    "doc = nlp(\"I have an acccount with chime since 2017\")\n",
    "print(doc.ents)\n",
    "\n",
    "patterns = [{\"label\": \"ORG\", \"pattern\": [{\"LOWER\": \"chime\"}]}]\n",
    "ruler = nlp.add_pipe(\"entity_ruler\")\n",
    "ruler.add_patterns(patterns)\n",
    "doc2 = nlp(\"I have an acccount with chime since 2017\")\n",
    "\n",
    "print(doc2.ents)\n",
    "print(doc2[5].ent_type_)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e3d8d3aa-e93f-4cb4-a2b3-88127fcb2d15",
   "metadata": {},
   "source": [
    "### Linguistic features + named entities"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "54816c20-3c12-44c4-b502-342ada6d7f9f",
   "metadata": {},
   "outputs": [],
   "source": [
    "#using POS tag information, dependency labels and named entities\n",
    "doc = nlp(\"Einstein lived in Zurich.\") \n",
    "[(ent.text, ent.label_) for ent in doc.ents]\n",
    "\n",
    "person_ents = [ent for ent in doc.ents if ent.label_ == \"PERSON\"]\n",
    "\n",
    "#We use head of the entity's last token \n",
    "#Check if the children of live contains prepositional attachment\n",
    "for person_ent in person_ents:\n",
    "    head = person_ent[-1].head\n",
    "    if head.lemma_ == \"live\":\n",
    "            preps = [token for token in head.children if token.dep_ == \"prep\"]\n",
    "            for prep in preps:\n",
    "                places = [token for token in prep.children if token.ent_type_ == \"GPE\"]\n",
    "            print({'person': person_ent, 'city': places,'past': head.tag_ == \"VBD\"}) \n",
    "\n",
    " "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f657af85-b344-4acc-8dde-58b73c3c53bc",
   "metadata": {},
   "source": [
    "# Chapter 5 Word Vectors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "eca83ac0-e56c-4ef9-a497-4bca15d39dfe",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Semantic similarity\n",
    "doc1 = nlp(\"I visited England.\")\n",
    "doc2 = nlp(\"I went to London.\") \n",
    "\n",
    "doc1[1:3].similarity(doc2[1:4]) # for 2 spans\n",
    "\n",
    "#comparing two token objects\n",
    "doc1[2].similarity(doc2[3]) \n",
    "\n",
    "#comparing two docs\n",
    "doc1.similarity(doc2) \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b265c7b6-5a2d-4b21-b998-015b32bf1f50",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Visualizing distance between vectors\n",
    "import matplotlib.pyplot as plt \n",
    "from sklearn.decomposition import PCA \n",
    "import numpy as np \n",
    "import spacy \n",
    "\n",
    "nlp = spacy.load(\"en_core_web_md\")\n",
    "\n",
    "vocab = nlp(\"cat dog tiger elephant bird monkey lion cheetah burger pizza food cheese wine salad noodles macaroni fruit vegetable\") \n",
    "\n",
    "words = [word.text for word in vocab] \n",
    "\n",
    "vecs = np.vstack([word.vector for word in vocab if word.has_vector]) \n",
    "\n",
    "pca = PCA(n_components=2) \n",
    "vecs_transformed = pca.fit_transform(vecs) \n",
    "\n",
    "plt.figure(figsize=(20,15)) \n",
    "plt.scatter(vecs_transformed[:,0], vecs_transformed[:,1]) \n",
    "\n",
    "for word, coord in zip(words, vecs_transformed):\n",
    "    x,y = coord\n",
    "    plt.text(x,y,word, size=15)\n",
    "\n",
    "plt.show() \n",
    " \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3e1dc0d8-b333-43af-904b-e8b0f5ade44b",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Categorizing text with semantic similarity\n",
    "sentences = nlp(\"I purchased a science fiction book last week. I loved everything related to this fragrance: light, floral and feminine... I purchased a bottle of wine.  \")\n",
    "key = nlp(\"perfume\") \n",
    "\n",
    "for sent in sentences.sents:\n",
    "    print(sent.similarity(key)) \n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f723fa60-9227-4e9b-8d36-411185acbd46",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Extracting key phrases\n",
    "doc = nlp(\"My beautiful and cute dog jumped over the fence\") \n",
    "\n",
    "doc.noun_chunks \n",
    "\n",
    "list(doc.noun_chunks)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1ef1a8bc-1184-488f-b56a-f09e7db88128",
   "metadata": {},
   "outputs": [],
   "source": [
    "for sent in sentences.sents: #iterate over sentences\n",
    "    nchunks = [nchunk.text for nchunk in sent.noun_chunks] #extract noun chunks\n",
    "    nchunk_doc = nlp(\" \".join(nchunks)) #joined noun chunks in the list into a string\n",
    "    print(nchunk_doc.similarity(key)) #compared the doc object noun chunks to key"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f5b27a4d-20f4-433e-b50a-1a931d5ac9f2",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Extracting and comparing named entities (proper nouns)\n",
    "doc1 = nlp(\"Google Search, often referred as Google, is the most popular search engine nowadays.\") \n",
    "doc2 = nlp(\"Microsoft Bing is another popular search engine. Microsoft is known by its star product Microsoft Windows, a popular operating system sold over the world.\") \n",
    "doc3 = nlp(\"The Dead Sea is the lowest lake in the world, located in the Jordan Valley of Israel. It is also the saltiest lake in the world.\") \n",
    "print(doc1.ents)\n",
    "print(doc2.ents) \n",
    "print(doc3.ents)\n",
    "\n",
    "#Compare similarity score\n",
    "ents1 = [ent.text for ent in doc1.ents] \n",
    "ents2 = [ent.text for ent in doc2.ents] \n",
    "ents3 = [ent.text for ent in doc3.ents] \n",
    "\n",
    "ents1 = nlp(\" \".join(ents1))\n",
    "ents2 = nlp(\" \".join(ents2))\n",
    "ents3 = nlp(\" \".join(ents3))\n",
    "print(ents1.similarity(ents2)) \n",
    "print(ents1.similarity(ents3)) \n",
    "print(ents2.similarity(ents3)) \n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "42c23836-437e-4616-931e-8d86f09b439e",
   "metadata": {},
   "source": [
    "# Chapter 6 Semanting Parsing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "67a0c575-1afa-4568-b05f-cca0d565c293",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9f41af69-fff2-41cb-bac7-e654973829d8",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Extracting named entities\n",
    "#Using dependency relations for intent recognition\n",
    "#Semantic similarity methods for semantic parsing \n",
    "\n",
    "dataset = pd.read_csv(\"/Users/tdubon/Documents/GitHub/Mastering-spaCy/Chapter06/data/atis_intents.csv\", header=None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "510c4ed7-5fb6-466f-aefc-a907fd300da6",
   "metadata": {},
   "outputs": [],
   "source": [
    "#first column  intent, second column user utterances.\n",
    "dataset.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e45c0cf2-61b6-401c-ab56-e1a751f77596",
   "metadata": {},
   "outputs": [],
   "source": [
    "for text in dataset[1].head():\n",
    "    print(text)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d8d5b26e-853b-4daa-8836-f7dacc86562c",
   "metadata": {},
   "outputs": [],
   "source": [
    "#frequency of utterances by intent \n",
    "grouped = dataset.groupby(0).size()\n",
    "print(grouped)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3a25f477-d970-4317-a0cb-854919e7d0f3",
   "metadata": {},
   "outputs": [],
   "source": [
    "#the frequency of each label\n",
    "from collections import Counter\n",
    "import spacy\n",
    "\n",
    "nlp = spacy.load(\"en_core_web_md\")\n",
    "corpus = open(\"/Users/tdubon/Documents/GitHub/Mastering-spaCy/Chapter06/data/atis_utterances.txt\", \"r\").read().split(\"\\n\")\n",
    "    \n",
    "all_ent_labels = [] \n",
    "for sentence in corpus: \n",
    "    doc = nlp(sentence.strip()) \n",
    "    ents = doc.ents \n",
    "    all_ent_labels += [ent.label_ for ent in ents] \n",
    "\n",
    "c = Counter(all_ent_labels) \n",
    "print(c) \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1710b383-126d-4582-8474-5912416141c2",
   "metadata": {},
   "outputs": [],
   "source": [
    "#observe many GPE labels - extracting location entities using Matcher\n",
    "# using ADP and GPE (part of speech and entity type) pattern\n",
    "import spacy\n",
    "from spacy.matcher import Matcher\n",
    "\n",
    "nlp = spacy.load(\"en_core_web_md\")\n",
    "matcher = Matcher(nlp.vocab)\n",
    "\n",
    "pattern = [{\"POS\": \"ADP\"}, {\"ENT_TYPE\": \"GPE\"}] #pattern matching 2 tokens - preposition (ADP = adpositin) and location entity\n",
    "matcher.add(\"prepositionLocation\", [pattern])\n",
    "doc = nlp(\"show me flights from denver to boston on tuesday\") #test it\n",
    "\n",
    "matches = matcher(doc)\n",
    "\n",
    "for mid, start, end in matches:\n",
    "    print(doc[start:end]) \n",
    "\n",
    "#inspect the text for variations in usage and whether the matcher catches these (such as abbreviations and others)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2d125549-5189-4351-a533-405255a927ef",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Extract the locations using the ORG entity label\n",
    "\n",
    "pattern = [{\"ENT_TYPE\": \"ORG\", \"OP\": \"+\" }] #to capture one or more occurences, we use the + operator\n",
    "matcher.add(\"AirlineName\", [pattern]) \n",
    "doc = nlp(\"what is the earliest united airlines flight flying from denver\")\n",
    "\n",
    "matches = matcher(doc) \n",
    "\n",
    "for mid,start,end in matches:\n",
    "    print(doc[start:end]) \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b726423c-fe5a-402f-a712-9f51fd1ab81c",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Extract the dates and times using ENT_TYPE: DATE and ENT_TYPE: TIME entity labels\n",
    "\n",
    "pattern = [{\"ENT_TYPE\": \"DATE\"}, {\"ENT_TYPE\": \"TIME\"}] \n",
    "matcher.add(\"DatesTimes\", [pattern]) \n",
    "doc = nlp(\"what is the earliest united airlines flight flying from denver on January 2022\")\n",
    "\n",
    "matches = matcher(doc) \n",
    "\n",
    "for mid,start,end in matches:\n",
    "    print(doc[start:end]) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ab542b66-c7e8-45c8-b939-3f262fc46089",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Extract abbreviations on type entities\n",
    "nlp = spacy.load(\"en_core_web_md\") \n",
    "\n",
    "pattern1 = [{\"TEXT\": {\"REGEX\": \"\\w{1,2}\\d{1,2}\"}}] #matches single token like d1, d10, ad1\n",
    "pattern2 = [{\"SHAPE\": { \"IN\": [\"x\", \"xx\"]}}, {\"SHAPE\": { \"IN\": [\"d\", \"dd\"]}}] #matches two token abbrev like ap 5, ap 57, a 5\n",
    "pattern3 = [{\"TEXT\": {\"IN\": [\"class\", \"code\", \"abbrev\", \"abbreviation\"]}}, {\"SHAPE\": { \"IN\": [\"x\", \"xx\"]}}] #matches two tokens\n",
    "pattern4 = [{\"POS\": \"NOUN\", \"SHAPE\": { \"IN\": [\"x\", \"xx\"]}}] #matches 1-2 letter words with NOUN POS tag\n",
    "\n",
    "matcher = Matcher(nlp.vocab) \n",
    "matcher.add(\"abbrevEnts\", [pattern1, pattern2, pattern3, pattern4]) \n",
    "\n",
    "sentences = [\n",
    "'what does restriction ap 57 mean',\n",
    "'what does the abbreviation co mean',\n",
    "'what does fare code qo mean',\n",
    "'what is the abbreviation d10',\n",
    "'what does code y mean',\n",
    "'what does the fare code f and fn mean',\n",
    "'what is booking class c'\n",
    "]\n",
    "\n",
    "\n",
    "for sent in sentences:\n",
    "   doc = nlp(sent)\n",
    "   matches = matcher(doc)\n",
    "   for mid, start, end in matches:\n",
    "     print(doc[start:end])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f06ba60f-79b7-478c-aed3-4bec760504a0",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Using dependency trees to extract entities\n",
    "# iterating the utterance dependency trees to find out whether the preposition to is syntactically related to the location entity\n",
    "import spacy \n",
    "\n",
    "nlp = spacy.load(\"en_core_web_md\") \n",
    "\n",
    "def reach_parent(source_token, dest_token):\n",
    "    source_token = source_token.head #retrieving the head, starting with source token\n",
    "    while source_token != dest_token: #iterate over the head of each token - stops when reaches the source or root\n",
    "        if source_token.head == source_token:\n",
    "            return None\n",
    "        source_token = source_token.head\n",
    "        return source_token\n",
    "    \n",
    "doc = nlp(\"I'm going to a conference in Munich.\")\n",
    "print(doc[-2])\n",
    "print(doc[3])   \n",
    "print(doc[-1]) \n",
    "\n",
    "print(reach_parent(doc[-2], doc[3]))\n",
    "print(reach_parent(doc[-1], doc[3]))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b4c424e6-a10d-482f-ae68-6388a0bc3530",
   "metadata": {},
   "source": [
    "### Dependency relations for intent recognition"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "91ab3c7f-fc8f-452e-98d8-e8c6e1957e7b",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Recognizing intent\n",
    "import spacy \n",
    "nlp = spacy.load(\"en_core_web_md\")\n",
    "\n",
    "doc = nlp(\"find a flight from washington to sf\")\n",
    "\n",
    "for token in doc:\n",
    "    if token.dep_ == \"dobj\": #search for dependency label dobj\n",
    "        print(token.head.text + token.text.capitalize()) #finding dobj's syntactic head\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "220a454d-fda0-4c06-8c26-e9e3227a48e0",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Extracting multiple intents with conjunction relation - intent: show flights\n",
    "#extracting the direct object and its conjuncts\n",
    "\n",
    "import spacy\n",
    "nlp = spacy.load(\"en_core_web_md\")\n",
    "doc = nlp(\"show all flights and fares from denver to san francisco\")\n",
    "\n",
    "for token in doc:\n",
    "    if token.dep_ == \"dobj\": #loop over all tokens to find the direct object\n",
    "        dobj = token.text\n",
    "        conj = [t.text for t in token.conjuncts]  #obtained conjuncts\n",
    "        verb = token.head.text #obtain transitive verb\n",
    "        #print(dobj, conj)\n",
    "\n",
    "print(verb, dobj, conj)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b19c3790-4fbc-450c-b076-fe91f56b58ce",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Extracting intent requiring deeper investigation of sentence structure - using list of helper verbs\n",
    "\n",
    "doc = nlp(\"i want to make a reservation for a flight\")\n",
    "\n",
    "dObj= None\n",
    "tVerb = None \n",
    "\n",
    "# Extract the direct object and its transitive verb\n",
    "for token in doc:\n",
    "    if token.dep_ == \"dobj\": #locate direct obj\n",
    "        dObj = token\n",
    "        tVerb = token.head #transitive verb\n",
    "\n",
    "# Extract the helper verb \n",
    "intentVerb = None \n",
    "verbList = [\"want\", \"like\", \"need\", \"order\"]\n",
    "\n",
    "if tVerb.text in verbList: #check transitive verb against verb list\n",
    "    intentVerb = tVerb\n",
    "else:\n",
    "    if tVerb.head.dep_ == \"ROOT\":\n",
    "        helperVerb = tVerb.head #check the main verb of the sentence, head of transitive root\n",
    "        \n",
    "\n",
    "# Extract the object of the intent \n",
    "intentObj = None\n",
    "objList = [\"flight\", \"meal\", \"booking\"] #list of the possible object words we want to extract\n",
    "\n",
    "if dObj.text in objList:\n",
    "    intentObj = dObj \n",
    "else:\n",
    "    for child in dObj.children: #if dobj not in list, check its synt. children\n",
    "        if child.dep_ == \"prep\": #check whether a preposition\n",
    "            intentObj = list(child.children)[0] #if yes, is the child a preposition of the dobj\n",
    "            break\n",
    "        elif child.dep_ == \"compound\": #if not prepos, check whether it is a compound\n",
    "            intentObj = child\n",
    "            break\n",
    "\n",
    "print(helperVerb.text + intentObj.text.capitalize()) \n",
    "\n",
    " "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3d7296aa-d601-408a-a5cc-6de0dc84b75c",
   "metadata": {},
   "outputs": [],
   "source": [
    "import spacy\n",
    "from spacy import displacy\n",
    "\n",
    "nlp = spacy.load(\"en_core_web_sm\")\n",
    "text = \"This study pointed to the assessment of the chemical composition (F, Ca, Mg, Na, K, Fe, Mn, Zn, Cu, Ni, Co, Cr, Cd, and carbohydrate) of different marine seaweeds (red, green, and brown) from the Egyptian Mediterranean Sea coast.\"\n",
    "doc = nlp(text)\n",
    "\n",
    "displacy.render(doc, style=\"ent\", jupyter=True)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c46277c0",
   "metadata": {},
   "source": [
    "### Semantic Parsing with List of Synonyms Comparison"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b816de0d-3cb5-4b82-a259-68c8dde55c66",
   "metadata": {},
   "outputs": [],
   "source": [
    "import spacy\n",
    "nlp = spacy.load(\"en_core_web_md\")\n",
    "\n",
    "#Include only the base form of the verbs and nouns\n",
    "#Each set includes the set of synonyms\n",
    "verbSynsets = [ (\"show\", \"list\"), (\"book\", \"make a reservation\", \"buy\", \"reserve\") ] \n",
    "\n",
    "objSynsets = [ (\"meal\", \"food\"), (\"aircraft\", \"airplane\", \"plane\") ] \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5061f317",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Comparing similarity of two utterances - Construct doc objects for each utterance \n",
    "\n",
    "doc = nlp(\"show me all aircrafts that cp uses\")\n",
    "\n",
    "doc2 = nlp(\"list all meals on my flight\") \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8ec0fd88",
   "metadata": {},
   "outputs": [],
   "source": [
    "[token.dep_ for token in doc2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "936b857f",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Extract transitive verb and direct object of each utterance\n",
    "\n",
    "for token in doc:\n",
    "    if token.dep_ == \"dobj\":\n",
    "        obj = token.lemma_\n",
    "        verb = token.head.lemma_\n",
    "        break \n",
    "        \n",
    "for token in doc2:\n",
    "    if token.dep_ == \"dobj\":\n",
    "        obj2 = token.lemma_\n",
    "        verb2 = token.head.lemma_\n",
    "        break \n",
    "        \n",
    "print(verb, obj)\n",
    "print(verb2, obj2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "26527504",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Identify the VERB synonyms in our list that appear in the user utterances\n",
    "vsyn = [syn for syn in verbSynsets if verb in syn]\n",
    "print(vsyn[0])\n",
    "print(verb2 in vsyn[0])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e0295cae",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Identify the synonyms for DOBJ in our list that appear in the user utterances\n",
    "osyn = [syn for syn in objSynsets if obj in syn]\n",
    "print(osyn[0])\n",
    "print(obj2 in osyn[0]) #object2 - meal - is not in list. Do not refer to same intent\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4e97ea8f",
   "metadata": {},
   "source": [
    "### Semantic Parsing using Word Vectors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "90fe9a7c",
   "metadata": {},
   "outputs": [],
   "source": [
    "import spacy\n",
    "nlp = spacy.load(\"en_core_web_md\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "59b2ad46",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Comparing similarity of two utterances - Construct doc objects for each utterance \n",
    "\n",
    "doc = nlp(\"show me all aircrafts that cp uses\")\n",
    "\n",
    "doc2 = nlp(\"list all meals on my flight\") "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2a4b8e7c",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Extract transitive verb and direct object of each utterance\n",
    "for token in doc:\n",
    "    if token.dep_ == \"dobj\":\n",
    "        obj = token\n",
    "        verb = token.head\n",
    "        break \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0e80e02d",
   "metadata": {},
   "outputs": [],
   "source": [
    "for token in doc2:\n",
    "    if token.dep_ == \"dobj\": #note this is object of the preposition\n",
    "        obj2 = token\n",
    "        verb2 = token.head\n",
    "        print(obj2, verb2)\n",
    "        break \n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fa481acc",
   "metadata": {},
   "outputs": [],
   "source": [
    "verb, obj\n",
    "verb2, obj2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "061682f2",
   "metadata": {},
   "outputs": [],
   "source": [
    "#calculate semantic similarity between the two direct objects\n",
    "obj.similarity(obj2) #low score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ba765f56",
   "metadata": {},
   "outputs": [],
   "source": [
    "#calculate similarity between the two verbs\n",
    "verb.similarity(verb2) #low score - not related"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "657afa53",
   "metadata": {},
   "source": [
    "### Intent recognition"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e3eecf84",
   "metadata": {},
   "outputs": [],
   "source": [
    "import spacy\n",
    "from spacy.matcher import Matcher\n",
    "nlp = spacy.load(\"en_core_web_md\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f43dcd57",
   "metadata": {},
   "outputs": [],
   "source": [
    "#step 1: extract entities\n",
    "matcher = Matcher(nlp.vocab)\n",
    "pattern = [{\"POS\": \"ADP\"}, {\"ENT_TYPE\": \"GPE\"}] #ADP for adpositions \n",
    "\n",
    "matcher.add(\"prepositionLocation\", [pattern]) \n",
    "\n",
    "# Location entities \n",
    "doc = nlp(\"show me flights from denver to philadelphia on tuesday\")\n",
    "matches = matcher(doc) \n",
    "\n",
    "for mid, start, end in matches:\n",
    "    print(doc[start:end]) \n",
    "\n",
    "#all entities\n",
    "ents = doc.ents\n",
    "print(ents)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a6ca7531",
   "metadata": {},
   "outputs": [],
   "source": [
    "#step 2: perform intent recognition\n",
    "\n",
    "for token in doc:\n",
    "    if token.dep_ == \"dobj\":\n",
    "        print(token.head.lemma_ + token.lemma_.capitalize())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9d298d0f",
   "metadata": {},
   "outputs": [],
   "source": [
    "{ \n",
    "    'utterance': 'show me flights from denver to philadelphia on tuesday', \n",
    "    'intent ': ' showFlight', \n",
    "    'entities': {\n",
    "        'date': 'tuesday',\n",
    "        'locations': {\n",
    "            'from': 'denver',\n",
    "            'to': 'philadelphia'\n",
    "        }\n",
    "    } \n",
    "} "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "240bf7e6",
   "metadata": {},
   "source": [
    "## Chapter 7"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4f28220d",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Are the entities and labels captured in the results\n",
    "import spacy\n",
    "nlp = spacy.load(\"en_core_web_md\")\n",
    "\n",
    "doc1 = nlp(\"navigate to my home\")\n",
    "print(doc1.ents)\n",
    "\n",
    "doc2 = nlp(\"navigate to Oxford Street\")\n",
    "print(doc2.ents)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "73aa52cd",
   "metadata": {},
   "outputs": [],
   "source": [
    "print(doc2.ents[0].label_)\n",
    "print(spacy.explain(\"ORG\")) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "91fa9793",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Home was not detected\n",
    "#Train NER component to recognize street names - add new entity type with EntityRuler\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e266b08a",
   "metadata": {},
   "source": [
    "### Creating an annotation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2cc02205",
   "metadata": {},
   "outputs": [],
   "source": [
    "## Prepare training data\n",
    "## Feed the raw text and annotations directly to spaCy\n",
    "\n",
    "import spacy\n",
    "from spacy.training import Example\n",
    "\n",
    "\n",
    "nlp = spacy.load(\"en_core_web_md\")\n",
    "doc = nlp(\"I will visit you in Munich\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "827556e7",
   "metadata": {},
   "outputs": [],
   "source": [
    "annotations = {\"entities\": [(20, 26, \"GPE\")]} \n",
    "\n",
    "#doc object and annotations fed into Example.from_dict()\n",
    "example_sent = Example.from_dict(doc, annotations) \n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6902f44f",
   "metadata": {},
   "outputs": [],
   "source": [
    "#add annotation to NER component\n",
    "\n",
    "#Disable other pipes - train only the one needed\n",
    "other_pipes = [pipe for pipe in nlp.pipe_names if pipe != 'ner']\n",
    "nlp.disable_pipes(*other_pipes) \n",
    "\n",
    "#Another way of writing this code is as follows: \n",
    "#other_pipes = [pipe for pipe in nlp.pipe_names if pipe != 'ner'] \n",
    "#with nlp.disable_pipes(*other_pipes):    \n",
    "    # training code goes here "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c6f2396e",
   "metadata": {},
   "source": [
    "### Creating a custom Named Entity Recognition model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "555eb761",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Train the NER model\n",
    "import random\n",
    "import spacy\n",
    "from spacy.training import Example\n",
    "from spacy.language import Language\n",
    "from pathlib import Path\n",
    "\n",
    "from __future__ import unicode_literals, print_function\n",
    "\n",
    "nlp = spacy.load(\"en_core_web_md\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "18cd3238",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define navigation domain training set sentences\n",
    "trainset = [(\"navigate home\", {\"entities\": [(9,13, \"GPE\")]}),\n",
    "            (\"navigate to office\", {\"entities\": [(12,18, \"GPE\")]}),\n",
    "            (\"navigate\", {\"entities\": []}), \n",
    "            (\"navigate to Oxford Street\", {\"entities\": [(12, 25, \"GPE\")]})] \n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "dc0ef662",
   "metadata": {},
   "outputs": [],
   "source": [
    "#setup variables\n",
    "model = None\n",
    "output_dir=Path(\"/Users/tdubon/spaCy_models\")\n",
    "n_iter=100"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8d6be714",
   "metadata": {},
   "outputs": [],
   "source": [
    "#load the model\n",
    "\n",
    "if model is not None:\n",
    "    nlp = spacy.load(model)  \n",
    "    print(\"Loaded model '%s'\" % model)\n",
    "else:\n",
    "    nlp = spacy.load('en_core_web_md')  \n",
    "\n",
    "\n",
    "#set up the pipeline\n",
    "if 'ner' not in nlp.pipe_names:\n",
    "    ner = nlp.create_pipe('ner')\n",
    "    nlp.add_pipe(ner, last=True)\n",
    "else:\n",
    "    ner = nlp.get_pipe('ner')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7aba7f34",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Disable other pipes\n",
    "other_pipes = [pipe for pipe in nlp.pipe_names if pipe != 'ner']\n",
    "\n",
    "#add labels\n",
    "#for _, annotations in trainset:\n",
    "#    for ent in annotations.get('entities'):\n",
    "#        ner.add_label(ent[2])\n",
    "\n",
    " #define component       \n",
    "def ner_component(data):\n",
    "    with nlp.disable_pipes(*other_pipes):\n",
    "        optimizer = nlp.create_optimizer()\n",
    "        for i in range(epochs):\n",
    "            random.shuffle(trainset)\n",
    "            for text, annotation in data:\n",
    "                doc = nlp.make_doc(text)\n",
    "                example = Example.from_dict(doc, annotation)\n",
    "                print(nlp.update([example], sgd=optimizer))\n",
    "\n",
    "ner_component(trainset)\n",
    "nlp.pipe_names"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d2322ae9",
   "metadata": {},
   "outputs": [],
   "source": [
    "#run model on data it was trained on\n",
    "for text, _ in trainset:\n",
    "    doc = nlp(text)\n",
    "    print('Entities', [(ent.text, ent.label_) for ent in doc.ents])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a6adc88e",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Evaluate updated NER model\n",
    "sents = [(\"navigate to my house\"), \n",
    "          (\"drive me home\"),\n",
    "          (\"navigate to Soho\"),\n",
    "          (\"navigate to my building\"), \n",
    "         (\"I watched a documentary about Lady Diana\")] \n",
    "\n",
    "for i in sents:\n",
    "           doc = nlp(i)\n",
    "           print(doc.ents)\n",
    "           print(doc.ents[0].label_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "cb71a5f1",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Saving custom model \n",
    "def save_model(output_dir):\n",
    "    if output_dir is not None:\n",
    "        output_dir = Path(output_dir)\n",
    "        if not output_dir.exists():\n",
    "            output_dir.mkdir()\n",
    "        nlp.to_disk(output_dir)\n",
    "        print(\"Saved model to\", output_dir)\n",
    "        \n",
    "save_model(output_dir)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b374f439",
   "metadata": {},
   "outputs": [],
   "source": [
    "### Loading custom model\n",
    "nlp = spacy.load(\"/Users/tdubon/spaCy_models\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5a8bff85",
   "metadata": {},
   "outputs": [],
   "source": [
    "#test\n",
    "sents = [(\"navigate to my house\"), \n",
    "          (\"drive me home\"),\n",
    "          (\"navigate to Soho\"),\n",
    "          (\"navigate to my building\"), \n",
    "         (\"I watched a documentary about Lady Diana\")]\n",
    "\n",
    "for i in sents:\n",
    "    doc = nlp(i)\n",
    "    print([(ent.text, ent.label_) for ent in doc.ents])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "111dca91",
   "metadata": {},
   "source": [
    "## Saving and Loading Custom Models Created using pipelines"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6731e368",
   "metadata": {},
   "outputs": [],
   "source": [
    "#need to look at**************Po\n",
    "import spacy\n",
    "\n",
    "nlp = spacy.load('en_core_web_md', disable=['ner']) #load pipeline w/o ner so we can use ours\n",
    "ner = nlp.create_pipe(\"ner\") #create pipeline component\n",
    "ner.from_disk(\"/Users/tdubon/spaCy_models/ner\") #load custom component\n",
    "nlp.add_pipe(ner, \"ner\") #add to pipeline\n",
    "print(nlp.meta['pipeline']) \n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6452b473",
   "metadata": {},
   "source": [
    "## Training an NER pipeline component from scratch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "496140e7",
   "metadata": {},
   "outputs": [],
   "source": [
    "import random\n",
    "import spacy\n",
    "import json\n",
    "from spacy.training import Example\n",
    "\n",
    "nlp = spacy.load(\"en_core_web_md\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "51b0def3",
   "metadata": {},
   "outputs": [],
   "source": [
    "with open(\"/Users/tdubon/Documents/GitHub/Mastering-spaCy/Chapter07/data/corona.json\") as f:\n",
    "    data = json.loads(f.read())\n",
    "\n",
    "TRAIN_DATA = [] \n",
    "for (text, annotation) in data:\n",
    "    new_anno = []\n",
    "    for anno in annotation[\"entities\"]:\n",
    "        st, end, label = anno\n",
    "        new_anno.append((st, end, label))\n",
    "    TRAIN_DATA.append((text, {\"entities\": new_anno})) \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "17f95523",
   "metadata": {},
   "outputs": [],
   "source": [
    "#training data\n",
    "train_set = [(\"Methylphenidate is effectively used in treating children with epilepsy and ADHD.\", \n",
    "              {\"entities\": [(0, 15, \"DRUG\"), (62, 70, \"DISEASE\"), (75, 79, \"DISEASE\")]}),\n",
    "             (\"Patients were followed up for 6 months.\", {\"entities\": []}),\n",
    "             (\"Antichlamydial antibiotics may be useful for curing coronary-artery disease.\", {\"entities\": [(0, 26, \"DRUG\"), (52, 75, \"DIS\")]}) ] \n",
    "\n",
    "#entities to recognize\n",
    "entities = [\"DIS\", \"DRUG\"]\n",
    "\n",
    "#define the labels we need NER to recognize\n",
    "labels = ['Pathogen', 'MedicalCondition', 'Medicine']\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b68cb62a",
   "metadata": {},
   "outputs": [],
   "source": [
    "#create a blank model\n",
    "nlp = spacy.blank(\"en\")\n",
    "ner = nlp.add_pipe(\"ner\")\n",
    "print(ner)\n",
    "print(nlp.meta)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "676e9151",
   "metadata": {},
   "outputs": [],
   "source": [
    "#add defined entity labels to model\n",
    "\n",
    "for ent in labels:\n",
    "    ner.add_label(ent)\n",
    "    print(ner.labels) \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c2f74786",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "\n",
    "#Disable other pipes\n",
    "other_pipes = [pipe for pipe in nlp.pipe_names if pipe != 'ner']\n",
    "\n",
    "#define update component       \n",
    "with nlp.disable_pipes(*other_pipes):\n",
    "    optimizer = nlp.begin_training()\n",
    "    for i in range(100):\n",
    "        random.shuffle(TRAIN_DATA)\n",
    "        for text, annotation in TRAIN_DATA:\n",
    "            doc = nlp.make_doc(text)\n",
    "            example = Example.from_dict(doc, annotation)\n",
    "            nlp.update([example], sgd=optimizer)\n",
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2b292f04",
   "metadata": {},
   "outputs": [],
   "source": [
    "nlp.pipe_names"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e66cdd04",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Test component containing new training data\n",
    "doc = nlp(\"I had a coronary disease\")\n",
    "print(f\"{doc.ents}, {doc.ents[0].label_}\") "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d7123d17",
   "metadata": {},
   "outputs": [],
   "source": [
    "from spacy import displacy \n",
    "\n",
    "doc = nlp(\"I had a coronary disease.\") \n",
    "displacy.render(doc, style=\"ent\") \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "53ce4d93",
   "metadata": {},
   "outputs": [],
   "source": [
    "doc2 = nlp(\"Pathogenic bacteria contribute to other globally important diseases, such as pneumonia, which can be caused by bacteria such as Streptococcus and Pseudomonas, and foodborne illnesses, which can be caused by bacteria such as Shigella, Campylobacter, and Salmonella.Pathogenic bacteria also cause infections such as tetanus, typhoid fever, diphtheria, syphilis, and leprosy. Pathogenic bacteria are also the cause of high infant mortality rates in developing countries.\") \n",
    "displacy.render(doc2, style=\"ent\") \n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e01d83ad",
   "metadata": {},
   "source": [
    "# Chapter 8"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e1746abc",
   "metadata": {},
   "source": [
    "## Training a text classifier - using TextCategorizer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "fd807a2f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<spacy.pipeline.textcat.TextCategorizer object at 0x7fd3a1206880>\n"
     ]
    }
   ],
   "source": [
    "#import BINARY classifier\n",
    "import spacy\n",
    "import random\n",
    "from spacy.training import Example\n",
    "from spacy.pipeline.textcat import DEFAULT_SINGLE_TEXTCAT_MODEL\n",
    "\n",
    "nlp = spacy.load(\"en_core_web_md\")\n",
    "\n",
    "#configure the TextCategorizer\n",
    "config = {\n",
    "    \"threshold\": 0.5,\n",
    "    \"model\": DEFAULT_SINGLE_TEXTCAT_MODEL }\n",
    "\n",
    "textcat = nlp.add_pipe(\"textcat\", config=config) \n",
    "\n",
    "print(textcat)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "14b39b92",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_data = [(\"I loved this product, very easy to use.\", {\"cats\": {\"POS\": True}}),\n",
    "              (\"I'll definitely purchase again. I recommend this product.\", {\"cats\": {\"POS\": True}}),\n",
    "              (\"This is the best product ever. I loved the scent and the feel. Will buy again.\", {\"cats\": {\"POS\": True}}),\n",
    "              (\"Disappointed. This product didn't work for me at all\", {\"cats\": {\"NEG\": True}}),\n",
    "              (\"I hated the scent. Won't buy again\", {\"cats\": {\"NEG\": True}}),\n",
    "              (\"Truly horrible product. Very few amount of product for a high price. Don't recommend.\", {\"cats\": {\"NEG\": True}}) ] \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "ad07d86b",
   "metadata": {},
   "outputs": [],
   "source": [
    "textcat.add_label(\"POS\")\n",
    "textcat.add_label(\"NEG\")\n",
    "    \n",
    "train_examples = [Example.from_dict(nlp.make_doc(text), label) for text,label in train_data] \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "e422d210",
   "metadata": {},
   "outputs": [],
   "source": [
    "textcat.initialize(lambda: train_examples, nlp=nlp)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "20f853be",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'textcat': 0.25}\n",
      "{'textcat': 0.2549639642238617}\n",
      "{'textcat': 0.2512111961841583}\n",
      "{'textcat': 0.25617092847824097}\n",
      "{'textcat': 0.23683953285217285}\n",
      "{'textcat': 0.27748462557792664}\n",
      "{'textcat': 0.2318648397922516}\n",
      "{'textcat': 0.21884527802467346}\n",
      "{'textcat': 0.26352161169052124}\n",
      "{'textcat': 0.21494124829769135}\n",
      "{'textcat': 0.25685396790504456}\n",
      "{'textcat': 0.2165045589208603}\n",
      "{'textcat': 0.2051442712545395}\n",
      "{'textcat': 0.2490426003932953}\n",
      "{'textcat': 0.2050033062696457}\n",
      "{'textcat': 0.22427338361740112}\n",
      "{'textcat': 0.1996849924325943}\n",
      "{'textcat': 0.21593083441257477}\n",
      "{'textcat': 0.19781598448753357}\n",
      "{'textcat': 0.2628738284111023}\n",
      "{'textcat': 0.14720141887664795}\n",
      "{'textcat': 0.14142610132694244}\n",
      "{'textcat': 0.27815133333206177}\n",
      "{'textcat': 0.2532503604888916}\n",
      "{'textcat': 0.11012145131826401}\n",
      "{'textcat': 0.10882793366909027}\n",
      "{'textcat': 0.2786612808704376}\n",
      "{'textcat': 0.30029404163360596}\n",
      "{'textcat': 0.09603079408407211}\n",
      "{'textcat': 0.33497944474220276}\n",
      "{'textcat': 0.2900579571723938}\n",
      "{'textcat': 0.10246645659208298}\n",
      "{'textcat': 0.24596786499023438}\n",
      "{'textcat': 0.1103532686829567}\n",
      "{'textcat': 0.2691419720649719}\n",
      "{'textcat': 0.12178987264633179}\n",
      "{'textcat': 0.10654643923044205}\n",
      "{'textcat': 0.27247002720832825}\n",
      "{'textcat': 0.27594974637031555}\n",
      "{'textcat': 0.09476035833358765}\n",
      "{'textcat': 0.09388066828250885}\n",
      "{'textcat': 0.2525210380554199}\n",
      "{'textcat': 0.08291522413492203}\n",
      "{'textcat': 0.27901265025138855}\n",
      "{'textcat': 0.2806483209133148}\n",
      "{'textcat': 0.08252662420272827}\n",
      "{'textcat': 0.08213651925325394}\n",
      "{'textcat': 0.2493847757577896}\n",
      "{'textcat': 0.24711936712265015}\n",
      "{'textcat': 0.08126690983772278}\n",
      "{'textcat': 0.07996674627065659}\n",
      "{'textcat': 0.07388274371623993}\n",
      "{'textcat': 0.3002031743526459}\n",
      "{'textcat': 0.29869765043258667}\n",
      "{'textcat': 0.27385005354881287}\n",
      "{'textcat': 0.2681586742401123}\n",
      "{'textcat': 0.06986071914434433}\n",
      "{'textcat': 0.08423873037099838}\n",
      "{'textcat': 0.07673240453004837}\n",
      "{'textcat': 0.25537487864494324}\n",
      "{'textcat': 0.06771593540906906}\n",
      "{'textcat': 0.25064677000045776}\n",
      "{'textcat': 0.2706637382507324}\n",
      "{'textcat': 0.256535142660141}\n",
      "{'textcat': 0.06714276969432831}\n",
      "{'textcat': 0.07778928428888321}\n",
      "{'textcat': 0.23293501138687134}\n",
      "{'textcat': 0.2338370382785797}\n",
      "{'textcat': 0.19864672422409058}\n",
      "{'textcat': 0.09699635207653046}\n",
      "{'textcat': 0.10439182072877884}\n",
      "{'textcat': 0.08647765964269638}\n",
      "{'textcat': 0.18809214234352112}\n",
      "{'textcat': 0.19814404845237732}\n",
      "{'textcat': 0.07614228874444962}\n",
      "{'textcat': 0.07749856263399124}\n",
      "{'textcat': 0.2369108349084854}\n",
      "{'textcat': 0.06287407130002975}\n",
      "{'textcat': 0.23643770813941956}\n",
      "{'textcat': 0.06420768052339554}\n",
      "{'textcat': 0.06061606481671333}\n",
      "{'textcat': 0.2579091787338257}\n",
      "{'textcat': 0.25403866171836853}\n",
      "{'textcat': 0.04803056642413139}\n",
      "{'textcat': 0.05375996232032776}\n",
      "{'textcat': 0.2527181804180145}\n",
      "{'textcat': 0.2511889636516571}\n",
      "{'textcat': 0.043591175228357315}\n",
      "{'textcat': 0.2561004161834717}\n",
      "{'textcat': 0.0425189808011055}\n",
      "{'textcat': 0.26279890537261963}\n",
      "{'textcat': 0.04069115221500397}\n",
      "{'textcat': 0.041983507573604584}\n",
      "{'textcat': 0.28144410252571106}\n",
      "{'textcat': 0.03548359125852585}\n",
      "{'textcat': 0.3275816738605499}\n",
      "{'textcat': 0.028371606022119522}\n",
      "{'textcat': 0.31586354970932007}\n",
      "{'textcat': 0.03148467466235161}\n",
      "{'textcat': 0.326810747385025}\n",
      "{'textcat': 0.03087560459971428}\n",
      "{'textcat': 0.2932112514972687}\n",
      "{'textcat': 0.3066202402114868}\n",
      "{'textcat': 0.33001720905303955}\n",
      "{'textcat': 0.30682066082954407}\n",
      "{'textcat': 0.044587038457393646}\n",
      "{'textcat': 0.04657663032412529}\n",
      "{'textcat': 0.04551272839307785}\n",
      "{'textcat': 0.04023783653974533}\n",
      "{'textcat': 0.27642834186553955}\n",
      "{'textcat': 0.2794535458087921}\n",
      "{'textcat': 0.028120428323745728}\n",
      "{'textcat': 0.4072722792625427}\n",
      "{'textcat': 0.03253287449479103}\n",
      "{'textcat': 0.27713483572006226}\n",
      "{'textcat': 0.26154860854148865}\n",
      "{'textcat': 0.04532147943973541}\n",
      "{'textcat': 0.04387292638421059}\n",
      "{'textcat': 0.2168249487876892}\n",
      "{'textcat': 0.03728972002863884}\n"
     ]
    }
   ],
   "source": [
    "#Define training example\n",
    "\n",
    "epochs = 20\n",
    "\n",
    "#Disable other pipe components & define training loop to incorporate statistical information\n",
    "\n",
    "with nlp.select_pipes(enable=\"textcat\"):\n",
    "    optimizer = nlp.resume_training() #Creates optimizer object\n",
    "    for i in range(epochs):\n",
    "        random.shuffle(train_data)\n",
    "        for text, label in train_data:\n",
    "            doc = nlp.make_doc(text)\n",
    "            example = Example.from_dict(doc, label) \n",
    "            print(nlp.update([example], sgd=optimizer))\n",
    "            \n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "116fa5c2",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'POS': 0.7234081625938416, 'NEG': 0.27659186720848083}\n"
     ]
    }
   ],
   "source": [
    "#test\n",
    "doc2 = nlp(\"This product sucks\")\n",
    "\n",
    "print(doc2.cats)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "9672da14",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'POS': 0.7797467708587646, 'NEG': 0.22025315463542938}\n"
     ]
    }
   ],
   "source": [
    "doc3 = nlp(\"This product is great\")\n",
    "print(doc3.cats)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "683b1008",
   "metadata": {},
   "source": [
    "## Multiclass Classification"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "bd989252",
   "metadata": {},
   "outputs": [],
   "source": [
    "## Training TextCategorizer for multilabel classification\n",
    "train_data = [(\"It's the perfect movie for a Sunday evening.\", {\"cats\": {\"SUNDAY_EVENING\": True}}),\n",
    "              (\"Very good thriller\", {\"cats\": {\"THRILLER\": True}}),\n",
    "              (\"A great movie for the kids and all the family\", {\"cats\": {\"FAMILY\": True}}),\n",
    "              (\"An ideal movie for Sunday night with all the family. My kids loved the movie.\", {\"cats\": {\"FAMILY\": True, \"SUNDAY_EVENING\":True}}),\n",
    "              (\"A perfect thriller for all the family. No violence, no drugs, pure action.\", {\"cats\": {\"FAMILY\": True, \"THRILLER\": True}})] \n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "5571c9ea",
   "metadata": {},
   "outputs": [],
   "source": [
    "import random\n",
    "import spacy\n",
    "from spacy.training import Example\n",
    "from spacy.pipeline.textcat_multilabel import DEFAULT_MULTI_TEXTCAT_MODEL\n",
    "\n",
    "nlp = spacy.load(\"en_core_web_md\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "69405cfc",
   "metadata": {},
   "outputs": [],
   "source": [
    "#setup configuration and add the respective pipe\n",
    "\n",
    "config = {\"threshold\": 0.5, \"model\": DEFAULT_MULTI_TEXTCAT_MODEL}\n",
    "\n",
    "textcat = nlp.add_pipe(\"textcat_multilabel\", config=config) \n",
    " "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "651d8af0",
   "metadata": {},
   "outputs": [],
   "source": [
    "labels = [\"FAMILY\", \"THRILLER\", \"SUNDAY_EVENING\"]\n",
    "\n",
    "for label in labels:\n",
    "    textcat.add_label(label)\n",
    "    \n",
    "train_examples = [Example.from_dict(nlp.make_doc(text), label) for text,label in train_data] \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "0ab00be5",
   "metadata": {},
   "outputs": [],
   "source": [
    "textcat.initialize(lambda: train_examples, nlp=nlp)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "732341c9",
   "metadata": {},
   "outputs": [],
   "source": [
    "epochs=20 \n",
    "\n",
    "with nlp.select_pipes(enable=\"textcat_multilabel\"):\n",
    "    optimizer = nlp.resume_training()\n",
    "    for i in range(epochs):\n",
    "        random.shuffle(train_data)\n",
    "        for text, label in train_data:\n",
    "            doc = nlp.make_doc(text)\n",
    "            example = Example.from_dict(doc, label)\n",
    "            nlp.update([example], sgd=optimizer) \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "5069d3df",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'FAMILY': 0.7342678904533386,\n",
       " 'THRILLER': 0.43685072660446167,\n",
       " 'SUNDAY_EVENING': 0.8150126338005066}"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "doc2 = nlp(\"Definitely in my Sunday movie night list\")\n",
    "doc2.cats"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c3225315",
   "metadata": {},
   "source": [
    "## Fine Food Reviews Sentiment Analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "fca2af85",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "c8a39b42",
   "metadata": {},
   "outputs": [],
   "source": [
    "reviews_df = pd.read_csv(\"/Users/tdubon/Documents/GitHub/Mastering-spaCy/Chapter08/data/Reviews.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "621471ff",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(3999, 10)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "reviews_df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "afc28b80",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Text</th>\n",
       "      <th>Score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>To be fair only one of my twins got gas from t...</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Just recieved our first order of these (they d...</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>This is a fantastic Chai Masala. I am very pic...</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Flavorful and has added nutrition!  You use le...</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>I thought this soup tasted good. I liked the t...</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                Text  Score\n",
       "0  To be fair only one of my twins got gas from t...      1\n",
       "1  Just recieved our first order of these (they d...      5\n",
       "2  This is a fantastic Chai Masala. I am very pic...      5\n",
       "3  Flavorful and has added nutrition!  You use le...      5\n",
       "4  I thought this soup tasted good. I liked the t...      3"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "reviews_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "d7bec191",
   "metadata": {},
   "outputs": [],
   "source": [
    "reviews_df = reviews_df[[\"Text\", \"Score\"]].dropna()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "205f9f32",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "ax=reviews_df.Score.value_counts().plot(kind='bar', colormap='Paired') \n",
    "plt.show() "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "196fd63f",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/var/folders/zn/m0vy_yh17lb67mrm0jhh0ty40000gn/T/ipykernel_38050/362032454.py:1: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  reviews_df.Score[reviews_df.Score <= 3] = 0\n",
      "/var/folders/zn/m0vy_yh17lb67mrm0jhh0ty40000gn/T/ipykernel_38050/362032454.py:2: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  reviews_df.Score[reviews_df.Score >= 4] = 1\n"
     ]
    }
   ],
   "source": [
    "reviews_df.Score[reviews_df.Score <= 3] = 0\n",
    "reviews_df.Score[reviews_df.Score >= 4] = 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "e1b14e0e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "ax=reviews_df.Score.value_counts().plot(kind='bar', colormap='Paired') \n",
    "plt.show() "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "bdbab4d8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Text</th>\n",
       "      <th>Score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>To be fair only one of my twins got gas from t...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Just recieved our first order of these (they d...</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>This is a fantastic Chai Masala. I am very pic...</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Flavorful and has added nutrition!  You use le...</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>I thought this soup tasted good. I liked the t...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                Text  Score\n",
       "0  To be fair only one of my twins got gas from t...      0\n",
       "1  Just recieved our first order of these (they d...      1\n",
       "2  This is a fantastic Chai Masala. I am very pic...      1\n",
       "3  Flavorful and has added nutrition!  You use le...      1\n",
       "4  I thought this soup tasted good. I liked the t...      0"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "reviews_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4f965167",
   "metadata": {},
   "source": [
    "## Training a Text Classifier - predict positive/negative"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "7cc4d78c",
   "metadata": {},
   "outputs": [],
   "source": [
    "import spacy\n",
    "import random\n",
    "from spacy.training import Example\n",
    "from spacy.pipeline.textcat_multilabel import DEFAULT_MULTI_TEXTCAT_MODEL \n",
    "nlp = spacy.load(\"en_core_web_md\") \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "978a975f",
   "metadata": {},
   "outputs": [],
   "source": [
    "#setup configuration and add multilabel component to pipeline\n",
    "config = { \n",
    "\n",
    "   \"threshold\": 0.5, \n",
    "\n",
    "   \"model\": DEFAULT_MULTI_TEXTCAT_MODEL \n",
    "\n",
    "} \n",
    "\n",
    "textcat = nlp.add_pipe(\"textcat_multilabel\", config=config) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "21d26432",
   "metadata": {},
   "outputs": [],
   "source": [
    "#converting format that can be taken in by spaCy\n",
    "train_examples = []\n",
    "\n",
    "for index, row in reviews_df.iterrows():\n",
    "    text = row[\"Text\"]\n",
    "    rating = row[\"Score\"]\n",
    "    label = {\"POS\": True, \"NEG\": False} if rating == 1 else {\"NEG\": True, \"POS\": False}\n",
    "    train_examples.append(Example.from_dict(nlp.make_doc(text), {\"cats\": label}))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "e28511d8",
   "metadata": {},
   "outputs": [],
   "source": [
    "#setup labels and initialize textcat\n",
    "textcat.add_label(\"POS\") \n",
    "textcat.add_label(\"NEG\") \n",
    "textcat.initialize(lambda: train_examples, nlp=nlp) \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "26caa44a",
   "metadata": {},
   "outputs": [],
   "source": [
    "#update nlp with statistical data\n",
    "epochs = 2\n",
    "\n",
    "with nlp.select_pipes(enable=\"textcat_multilabel\"): \n",
    "    optimizer = nlp.resume_training()\n",
    "    for i in range(epochs): \n",
    "        random.shuffle(train_examples)\n",
    "        for example in train_examples:\n",
    "            nlp.update([example], sgd=optimizer) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "9264be05",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'POS': 0.9839656352996826, 'NEG': 0.02120601199567318}"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "doc2 = nlp(\"This is the best food I ever ate\")\n",
    "doc2.cats"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "946276ee",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'POS': 0.2525370419025421, 'NEG': 0.7667326927185059}"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "doc3 = nlp(\"This food is so bad\")\n",
    "doc3.cats"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6dae1420",
   "metadata": {},
   "source": [
    "## Text classification using Keras\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "f73ffedb",
   "metadata": {},
   "outputs": [],
   "source": [
    "data = [ \"Tomorrow I will visit the hospital.\",\n",
    "        \"Yesterday I took a flight to Athens.\", \"Sally visited Harry and his dog.\" ] \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "ccc3fc2f",
   "metadata": {},
   "outputs": [],
   "source": [
    "import spacy\n",
    "nlp = spacy.load(\"en_core_web_md\")\n",
    "\n",
    "#convert sentences to doc objects retrieve tokens\n",
    "sentences = [[token.text for token in nlp(sentence)] for sentence in data] \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "8c76cf2e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['Tomorrow', 'I', 'will', 'visit', 'the', 'hospital', '.']\n",
      "['Yesterday', 'I', 'took', 'a', 'flight', 'to', 'Athens', '.']\n",
      "['Sally', 'visited', 'Harry', 'and', 'his', 'dog', '.']\n"
     ]
    }
   ],
   "source": [
    "#inspect conversion\n",
    "for sentence in sentences:\n",
    "    print(sentence)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e7447dd2",
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "# may be necessary to install\n",
    "#!pip install tensorflow"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "a83770af",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<keras_preprocessing.text.Tokenizer at 0x7fd380e6eca0>"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#using Tokenizer object to turn word sequences to word-ID sequences\n",
    "import tensorflow as tf\n",
    "from tensorflow.keras.preprocessing.text import Tokenizer\n",
    "\n",
    "#create tokenizer object\n",
    "tokenizer = Tokenizer(lower=True) #lower all words\n",
    "tokenizer.fit_on_texts(data) #to fit on sequence of words\n",
    "tokenizer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "7865ea34",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'i': 1,\n",
       " 'tomorrow': 2,\n",
       " 'will': 3,\n",
       " 'visit': 4,\n",
       " 'the': 5,\n",
       " 'hospital': 6,\n",
       " 'yesterday': 7,\n",
       " 'took': 8,\n",
       " 'a': 9,\n",
       " 'flight': 10,\n",
       " 'to': 11,\n",
       " 'athens': 12,\n",
       " 'sally': 13,\n",
       " 'visited': 14,\n",
       " 'harry': 15,\n",
       " 'and': 16,\n",
       " 'his': 17,\n",
       " 'dog': 18}"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#print results\n",
    "tokenizer.word_index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "0a578fdf",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[[6], [8]]"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#retrieve the word-IDs\n",
    "tokenizer.texts_to_sequences([\"hospital\", \"took\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "4d54515c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['will tomorrow i', 'the hospital flight']"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#sequences_to_texts\n",
    "tokenizer.sequences_to_texts([[3,2,1], [5,6,10]])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "0aa2f188",
   "metadata": {},
   "outputs": [],
   "source": [
    "#zero padding - zeros added to make vectors all the same length\n",
    "#transform sentences into a sequence of word-ids\n",
    "from tensorflow.keras.preprocessing.sequence import pad_sequences"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "id": "07b14a04",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 7,  0,  0,  0],\n",
       "       [ 8,  1,  0,  0],\n",
       "       [ 9, 11, 12, 14]], dtype=int32)"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#sequences - list of list that correspond to word sequences\n",
    "sequences = [[7], [8,1], [9,11,12,14]] \n",
    "\n",
    "MAX_LEN=4 \n",
    "\n",
    "#pad the sequences\n",
    "tf.keras.preprocessing.sequence.pad_sequences(sequences, MAX_LEN, padding=\"post\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "94383ded",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 0,  0,  0,  7],\n",
       "       [ 0,  0,  8,  1],\n",
       "       [ 9, 11, 12, 14]], dtype=int32)"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tf.keras.preprocessing.sequence.pad_sequences(sequences, MAX_LEN, padding=\"pre\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a3da2ff1",
   "metadata": {},
   "source": [
    "## Keras implementation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "id": "10669b25",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "id": "5664ec68",
   "metadata": {},
   "outputs": [],
   "source": [
    "import spacy\n",
    "nlp = spacy.load(\"en_core_web_md\") "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "id": "3c017f94",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(3999, 10)"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#import data set\n",
    "\n",
    "reviews_df = pd.read_csv(\"/Users/tdubon/Documents/GitHub/Mastering-spaCy/Chapter08/data/Reviews.csv\")\n",
    "reviews_df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "id": "9e0c7570",
   "metadata": {},
   "outputs": [],
   "source": [
    "#extract review text and review label from each dataset row to append to python list\n",
    "train_examples = []\n",
    "labels = []\n",
    "\n",
    "for index, row in reviews_df.iterrows():\n",
    "    text = row[\"Text\"]\n",
    "    rating = row[\"Score\"]\n",
    "    labels.append(rating)\n",
    "    tokens = [token.text for token in nlp(text)]\n",
    "    train_examples.append(tokens)    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "id": "76747243",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['To',\n",
       " 'be',\n",
       " 'fair',\n",
       " 'only',\n",
       " 'one',\n",
       " 'of',\n",
       " 'my',\n",
       " 'twins',\n",
       " 'got',\n",
       " 'gas',\n",
       " 'from',\n",
       " 'this',\n",
       " 'but',\n",
       " 'it',\n",
       " 'was',\n",
       " 'horrible',\n",
       " '.',\n",
       " 'Up',\n",
       " 'all',\n",
       " 'night',\n",
       " 'screaming',\n",
       " 'from',\n",
       " 'gas',\n",
       " 'pains',\n",
       " '.',\n",
       " 'Garbanzo',\n",
       " 'beans',\n",
       " 'are',\n",
       " 'not',\n",
       " 'an',\n",
       " 'ideal',\n",
       " 'food',\n",
       " 'for',\n",
       " 'young',\n",
       " 'babies',\n",
       " '.']"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_examples[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "id": "b679015d",
   "metadata": {},
   "outputs": [],
   "source": [
    "#tokenize the data and create vocabulary\n",
    "from tensorflow.keras.preprocessing.text import Tokenizer\n",
    "from tensorflow.keras.preprocessing.sequence import pad_sequences\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "id": "3474b8a6",
   "metadata": {},
   "outputs": [],
   "source": [
    "#apply sequences to list containing list of words\n",
    "tokenizer = Tokenizer(lower=True)\n",
    "tokenizer.fit_on_texts(train_examples)\n",
    "\n",
    "sequences = tokenizer.texts_to_sequences(train_examples)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "id": "bbe57452",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[7,\n",
       " 40,\n",
       " 1489,\n",
       " 71,\n",
       " 48,\n",
       " 10,\n",
       " 15,\n",
       " 6075,\n",
       " 159,\n",
       " 1158,\n",
       " 54,\n",
       " 12,\n",
       " 19,\n",
       " 8,\n",
       " 24,\n",
       " 1187,\n",
       " 1,\n",
       " 69,\n",
       " 44,\n",
       " 661,\n",
       " 6076,\n",
       " 54,\n",
       " 1158,\n",
       " 3346,\n",
       " 1,\n",
       " 4895,\n",
       " 387,\n",
       " 22,\n",
       " 23,\n",
       " 76,\n",
       " 1927,\n",
       " 65,\n",
       " 13,\n",
       " 1667,\n",
       " 1928,\n",
       " 1]"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sequences[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "id": "37404848",
   "metadata": {},
   "outputs": [],
   "source": [
    "#apply padding to end of vectors\n",
    "MAX_LEN = 50\n",
    "\n",
    "X = pad_sequences(sequences, MAX_LEN, padding=\"post\")\n",
    "\n",
    "X = np.array(X)\n",
    "y = np.array(labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "id": "2ef44b4c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(3999, 1)"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y = y.reshape(y.shape[0] , 1)\n",
    "y.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "id": "d475c7d5",
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.models import Model\n",
    "from tensorflow.keras.layers import Input, LSTM, Dense, Embedding\n",
    "from tensorflow.keras import optimizers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "id": "13756525",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-01-21 07:59:42.202560: I tensorflow/core/platform/cpu_feature_guard.cc:151] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA\n",
      "To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.\n"
     ]
    }
   ],
   "source": [
    "#define the input layer\n",
    "sentence_input = Input(shape=(None,)) #none - asks keras to infer the shape\n",
    "\n",
    "#define the embedding layer, input dimension should be number of words in vocabulary +1\n",
    "embedding =  Embedding(input_dim = len(tokenizer.word_index)+1, \n",
    "                       output_dim = 100)(sentence_input)\n",
    "\n",
    "#feed word vectors to LSTM - units refer to hidden state dimension\n",
    "LSTM_layer = LSTM(units=256)(embedding)\n",
    "\n",
    "#convert output to one dimension using the sigmoid function (0 or 1 possible)\n",
    "output_dense = Dense(1, activation='sigmoid')(LSTM_layer)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "id": "9f99d9b6",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"model\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " input_1 (InputLayer)        [(None, None)]            0         \n",
      "                                                                 \n",
      " embedding (Embedding)       (None, None, 100)         1623900   \n",
      "                                                                 \n",
      " lstm (LSTM)                 (None, 256)               365568    \n",
      "                                                                 \n",
      " dense (Dense)               (None, 1)                 257       \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 1,989,725\n",
      "Trainable params: 1,989,725\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model = Model(inputs=[sentence_input],outputs=[output_dense])\n",
    "model.summary()\n",
    "\n",
    "#compile model with optimizer, loss function and eval. metric\n",
    "model.compile(optimizer=\"adam\", loss=\"binary_crossentropy\", metrics=[\"accuracy\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "id": "16a6439e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/5\n",
      "50/50 [==============================] - 16s 272ms/step - loss: -62.1481 - accuracy: 0.0907 - val_loss: -93.3616 - val_accuracy: 0.1075\n",
      "Epoch 2/5\n",
      "50/50 [==============================] - 14s 279ms/step - loss: -115.8135 - accuracy: 0.0916 - val_loss: -133.5630 - val_accuracy: 0.1075\n",
      "Epoch 3/5\n",
      "50/50 [==============================] - 13s 266ms/step - loss: -156.2205 - accuracy: 0.0916 - val_loss: -172.4322 - val_accuracy: 0.1075\n",
      "Epoch 4/5\n",
      "50/50 [==============================] - 13s 258ms/step - loss: -195.6194 - accuracy: 0.0916 - val_loss: -210.6899 - val_accuracy: 0.1075\n",
      "Epoch 5/5\n",
      "50/50 [==============================] - 13s 262ms/step - loss: -234.6112 - accuracy: 0.0916 - val_loss: -249.1014 - val_accuracy: 0.1075\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x7fd36a099b50>"
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#x is a list of training examples, y list of labels\n",
    "model.fit(x=X,\n",
    "          y=y,\n",
    "          batch_size=64,\n",
    "          epochs=5,\n",
    "          validation_split=0.2)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "67519fa6",
   "metadata": {},
   "source": [
    "# Chapter 9"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5151630a",
   "metadata": {},
   "source": [
    "## Modeling with transformers and BERT"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "79d4e027",
   "metadata": {},
   "outputs": [],
   "source": [
    "!pip install transformers"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "01aecb10",
   "metadata": {},
   "source": [
    "### Method 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "id": "a78e7474",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "05459e38e7914c50b3e31434e82c7413",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Downloading:   0%|          | 0.00/226k [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "dcda5df82a9b4504b6e26b2d2fbd900b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Downloading:   0%|          | 0.00/28.0 [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1ad64575cf2e4799af2f1595504385ff",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Downloading:   0%|          | 0.00/455k [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "230a9f9883a6466eb2ec1c44edbacff8",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Downloading:   0%|          | 0.00/570 [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['[CLS]', 'he', 'lived', 'characteristic', '##ally', 'idle', 'and', 'romantic', '.', '[SEP]']\n",
      "[101, 2002, 2973, 8281, 3973, 18373, 1998, 6298, 1012, 102]\n"
     ]
    }
   ],
   "source": [
    "#Tokenize\n",
    "from transformers import BertTokenizer\n",
    "\n",
    "#load pretraind model using .from_pretrained method\n",
    "btokenizer = BertTokenizer.from_pretrained('bert-base-uncased')\n",
    "\n",
    "#tokenize using .tokenize method\n",
    "sentence = \"He lived characteristically idle and romantic.\"\n",
    "sentence = \"[CLS] \" + sentence + \" [SEP]\"\n",
    "tokens = btokenizer.tokenize(sentence)\n",
    "\n",
    "print(tokens)\n",
    "\n",
    "#convert tokens to token ids using .convert_tokens_to_ids method\n",
    "ids = btokenizer.convert_tokens_to_ids(tokens)\n",
    "print(ids)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a5c60e7f",
   "metadata": {},
   "source": [
    "### Method 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "id": "03d44b7f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[101, 2002, 2973, 8281, 3973, 18373, 1998, 6298, 1012, 102]"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#previous code block (load pretrained model, tokenize, convert to token ids) can be implemented \n",
    "# + adding special tokens,  using encode method\n",
    "\n",
    "from transformers import BertTokenizer\n",
    "btokenizer = BertTokenizer.from_pretrained(\"bert-base-uncased\")\n",
    "sentence = \"He lived characteristically idle and romantic.\"\n",
    "\n",
    "ids = btokenizer.encode(sentence)\n",
    "ids"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d345601e",
   "metadata": {},
   "source": [
    "### Method 3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "id": "72688981",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Truncation was not explicitly activated but `max_length` is provided a specific value, please use `truncation=True` to explicitly truncate examples to max length. Defaulting to 'longest_first' truncation strategy. If you encode pairs of sequences (GLUE-style) with the tokenizer you can select this strategy more precisely by providing a specific strategy to `truncation`.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tf.Tensor([[  101  2002  2973  8281  3973 18373  1998  6298  1012   102     0     0]], shape=(1, 12), dtype=int32)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/tdubon/opt/anaconda3/lib/python3.9/site-packages/transformers/tokenization_utils_base.py:2226: FutureWarning: The `pad_to_max_length` argument is deprecated and will be removed in a future version, use `padding=True` or `padding='longest'` to pad to the longest sequence in the batch, or use `padding='max_length'` to pad to a max length. In this case, you can give a specific length with `max_length` (e.g. `max_length=45`) or leave max_length to None to pad to the maximal input size of the model (e.g. 512 for Bert).\n",
      "  warnings.warn(\n"
     ]
    }
   ],
   "source": [
    "#all steps, including padding, can be implemented using encode_plus\n",
    "from transformers import BertTokenizer\n",
    "\n",
    "btokenizer = BertTokenizer.from_pretrained('bert-base-uncased')\n",
    "\n",
    "sentence = \"He lived characteristically idle and romantic.\"\n",
    "\n",
    "encoded = btokenizer.encode_plus(text=sentence,\n",
    "                                 add_special_tokens=True,\n",
    "                                 max_length=12,\n",
    "                                 pad_to_max_length=True,\n",
    "                                 return_tensors=\"tf\" )\n",
    "\n",
    "token_ids = encoded[\"input_ids\"]\n",
    "print(token_ids) \n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b0a87a95",
   "metadata": {},
   "source": [
    "## Extract BERT embeddings"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "id": "837ebdea",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7bf1f5f0e9ff48c0b3b668956b293835",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Downloading:   0%|          | 0.00/511M [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Some layers from the model checkpoint at bert-base-uncased were not used when initializing TFBertModel: ['nsp___cls', 'mlm___cls']\n",
      "- This IS expected if you are initializing TFBertModel from the checkpoint of a model trained on another task or with another architecture (e.g. initializing a BertForSequenceClassification model from a BertForPreTraining model).\n",
      "- This IS NOT expected if you are initializing TFBertModel from the checkpoint of a model that you expect to be exactly identical (initializing a BertForSequenceClassification model from a BertForSequenceClassification model).\n",
      "All the layers of TFBertModel were initialized from the model checkpoint at bert-base-uncased.\n",
      "If your task is similar to the task the model of the checkpoint was trained on, you can already use TFBertModel for predictions without further training.\n",
      "Truncation was not explicitly activated but `max_length` is provided a specific value, please use `truncation=True` to explicitly truncate examples to max length. Defaulting to 'longest_first' truncation strategy. If you encode pairs of sequences (GLUE-style) with the tokenizer you can select this strategy more precisely by providing a specific strategy to `truncation`.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1, 10, 768)\n",
      "(1, 768)\n"
     ]
    }
   ],
   "source": [
    "\n",
    "\n",
    "from transformers import BertTokenizer, TFBertModel\n",
    "\n",
    "btokenizer = BertTokenizer.from_pretrained('bert-base-uncased')\n",
    "\n",
    "#Load bert-base-uncased model\n",
    "bmodel = TFBertModel.from_pretrained(\"bert-base-uncased\") \n",
    "\n",
    "sentence = \"He was idle.\"\n",
    "\n",
    "#use encode plus to transform input sentence to BERT\n",
    "encoded = btokenizer.encode_plus(\n",
    "        text=sentence,\n",
    "        add_special_tokens=True,\n",
    "        max_length=10,\n",
    "        pad_to_max_length=True,\n",
    "        return_attention_mask=True,\n",
    "        return_tensors=\"tf\"\n",
    ")\n",
    "\n",
    "#capture the tf.tensor results \n",
    "inputs = encoded[\"input_ids\"]\n",
    "\n",
    "#feed sentence to model, output is a \n",
    "#first tuple:  (batch size, \n",
    "                #sequence length - equivalent to max_length, \n",
    "                #hidden size - no of words x 768 dimensional vectors)\n",
    "#second tuple: (batch size, \n",
    "                #hidden size - word embedding of the [CLS] token)\n",
    "outputs = bmodel(inputs)\n",
    "\n",
    "print(outputs[0].shape)\n",
    "print(outputs[1].shape)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "88ba3d59",
   "metadata": {},
   "source": [
    "## BERT for text classification"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "id": "68651b55",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import tensorflow\n",
    "from tensorflow.keras.layers import Dense, Input\n",
    "from tensorflow.keras.models import Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "id": "55d916d8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(5572, 5)"
      ]
     },
     "execution_count": 88,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#import data set\n",
    "data = pd.read_csv(\"/Users/tdubon/Documents/GitHub/Mastering-spaCy/Chapter09/data/spam.csv\", encoding='ISO-8859-1')\n",
    "data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "id": "f6a51494",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>v1</th>\n",
       "      <th>v2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ham</td>\n",
       "      <td>Go until jurong point, crazy.. Available only ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ham</td>\n",
       "      <td>Ok lar... Joking wif u oni...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>spam</td>\n",
       "      <td>Free entry in 2 a wkly comp to win FA Cup fina...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>ham</td>\n",
       "      <td>U dun say so early hor... U c already then say...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>ham</td>\n",
       "      <td>Nah I don't think he goes to usf, he lives aro...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     v1                                                 v2\n",
       "0   ham  Go until jurong point, crazy.. Available only ...\n",
       "1   ham                      Ok lar... Joking wif u oni...\n",
       "2  spam  Free entry in 2 a wkly comp to win FA Cup fina...\n",
       "3   ham  U dun say so early hor... U c already then say...\n",
       "4   ham  Nah I don't think he goes to usf, he lives aro..."
      ]
     },
     "execution_count": 90,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.drop(data.columns[2:], axis=1, inplace=True)\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "id": "2a3b01ba",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>label</th>\n",
       "      <th>text</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ham</td>\n",
       "      <td>Go until jurong point, crazy.. Available only ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ham</td>\n",
       "      <td>Ok lar... Joking wif u oni...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>spam</td>\n",
       "      <td>Free entry in 2 a wkly comp to win FA Cup fina...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>ham</td>\n",
       "      <td>U dun say so early hor... U c already then say...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>ham</td>\n",
       "      <td>Nah I don't think he goes to usf, he lives aro...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  label                                               text\n",
       "0   ham  Go until jurong point, crazy.. Available only ...\n",
       "1   ham                      Ok lar... Joking wif u oni...\n",
       "2  spam  Free entry in 2 a wkly comp to win FA Cup fina...\n",
       "3   ham  U dun say so early hor... U c already then say...\n",
       "4   ham  Nah I don't think he goes to usf, he lives aro..."
      ]
     },
     "execution_count": 91,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.columns = [\"label\", \"text\"]\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "id": "deb68968",
   "metadata": {},
   "outputs": [],
   "source": [
    "data=data.dropna()                                                           \n",
    "data=data.reset_index(drop=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "id": "801d80a3",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Some layers from the model checkpoint at bert-base-uncased were not used when initializing TFBertModel: ['nsp___cls', 'mlm___cls']\n",
      "- This IS expected if you are initializing TFBertModel from the checkpoint of a model trained on another task or with another architecture (e.g. initializing a BertForSequenceClassification model from a BertForPreTraining model).\n",
      "- This IS NOT expected if you are initializing TFBertModel from the checkpoint of a model that you expect to be exactly identical (initializing a BertForSequenceClassification model from a BertForSequenceClassification model).\n",
      "All the layers of TFBertModel were initialized from the model checkpoint at bert-base-uncased.\n",
      "If your task is similar to the task the model of the checkpoint was trained on, you can already use TFBertModel for predictions without further training.\n"
     ]
    }
   ],
   "source": [
    "from transformers import BertTokenizer, TFBertModel, BertConfig, TFBertForSequenceClassification\n",
    "bert_tokenizer = BertTokenizer.from_pretrained(\"bert-base-uncased\")\n",
    "\n",
    "#Note that model starts with TF - refers to HuggingFace pre-trained models for tensorflow\n",
    "bmodel = TFBertModel.from_pretrained(\"bert-base-uncased\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "id": "befd5646",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>label</th>\n",
       "      <th>text</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>Go until jurong point, crazy.. Available only ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>Ok lar... Joking wif u oni...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>Free entry in 2 a wkly comp to win FA Cup fina...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>U dun say so early hor... U c already then say...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>Nah I don't think he goes to usf, he lives aro...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   label                                               text\n",
       "0      0  Go until jurong point, crazy.. Available only ...\n",
       "1      0                      Ok lar... Joking wif u oni...\n",
       "2      1  Free entry in 2 a wkly comp to win FA Cup fina...\n",
       "3      0  U dun say so early hor... U c already then say...\n",
       "4      0  Nah I don't think he goes to usf, he lives aro..."
      ]
     },
     "execution_count": 95,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data['label'] = data['label'].map({'ham':0,'spam':1})\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "id": "f2f04c5a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(5572, 5572)"
      ]
     },
     "execution_count": 96,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sentences=data['text']\n",
    "labels=data['label']\n",
    "len(sentences),len(labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "id": "c95a64b9",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Truncation was not explicitly activated but `max_length` is provided a specific value, please use `truncation=True` to explicitly truncate examples to max length. Defaulting to 'longest_first' truncation strategy. If you encode pairs of sequences (GLUE-style) with the tokenizer you can select this strategy more precisely by providing a specific strategy to `truncation`.\n",
      "/Users/tdubon/opt/anaconda3/lib/python3.9/site-packages/transformers/tokenization_utils_base.py:2226: FutureWarning: The `pad_to_max_length` argument is deprecated and will be removed in a future version, use `padding=True` or `padding='longest'` to pad to the longest sequence in the batch, or use `padding='max_length'` to pad to a max length. In this case, you can give a specific length with `max_length` (e.g. `max_length=45`) or leave max_length to None to pad to the maximal input size of the model (e.g. 512 for Bert).\n",
      "  warnings.warn(\n"
     ]
    }
   ],
   "source": [
    "#generate token ids\n",
    "\n",
    "input_ids=[]\n",
    "attention_masks=[]\n",
    "\n",
    "for sent in sentences:\n",
    "    bert_inp=bert_tokenizer.encode_plus(sent,add_special_tokens = True,max_length =64,pad_to_max_length = True,return_attention_mask = True)\n",
    "    input_ids.append(bert_inp['input_ids'])\n",
    "    attention_masks.append(bert_inp['attention_mask'])\n",
    "\n",
    "input_ids=np.asarray(input_ids)\n",
    "attention_masks=np.array(attention_masks)\n",
    "labels=np.array(labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "id": "7a684b9f",
   "metadata": {},
   "outputs": [],
   "source": [
    "#define input layers\n",
    "inputs = Input(shape=(64,), dtype=\"int32\")\n",
    "\n",
    "#feed inputs to model\n",
    "bert = bmodel(inputs)\n",
    "\n",
    "#extract the second output of BERT output\n",
    "bert = bert[1]\n",
    "\n",
    "#flatten to single layere with activation funct\n",
    "outputs = Dense(units=1, activation=\"sigmoid\")(bert)\n",
    "#\n",
    "model = Model(inputs, outputs)\n",
    "\n",
    "adam = tensorflow.keras.optimizers.Adam(learning_rate=2e-5,epsilon=1e-08)\n",
    "model.compile(loss=\"binary_crossentropy\", metrics=[\"accuracy\"], optimizer=adam)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "45cda449",
   "metadata": {},
   "outputs": [],
   "source": [
    "history=model.fit(input_ids,labels,batch_size=1,epochs=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "58818698",
   "metadata": {},
   "outputs": [],
   "source": [
    "bmodel.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "eb0911cc",
   "metadata": {},
   "source": [
    "## Transformer pipelines"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "329f0b00",
   "metadata": {},
   "outputs": [],
   "source": [
    "#using HuggingFace Transformer libraries - sentiment analysis\n",
    "\n",
    "#import the pipeline function\n",
    "from transformers import pipeline \n",
    "\n",
    "#load it\n",
    "nlp = pipeline(\"sentiment-analysis\")\n",
    "\n",
    "#sentences\n",
    "sent1 = \"I hate you so much right now.\"\n",
    "sent2 = \"I love fresh air and exercising.\" \n",
    "\n",
    "#\n",
    "result1 = nlp(sent1) \n",
    "result2 = nlp(sent2) \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4be5bde7",
   "metadata": {},
   "outputs": [],
   "source": [
    "from transformers import pipeline\n",
    "\n",
    "nlp = pipeline(\"question-answering\")\n",
    "res = nlp({'question': 'What is the name of this book?',\n",
    "           'context': \"I'll publish my new book Mastering spaCy soon.\" })\n",
    "print(res) \n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ec7df56b",
   "metadata": {},
   "source": [
    "## Transformers and spaCy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "8fa36200",
   "metadata": {},
   "outputs": [],
   "source": [
    "import spacy\n",
    "nlp = spacy.load(\"en_core_web_trf\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "b6f6404f",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/tdubon/opt/anaconda3/lib/python3.9/site-packages/torch/autocast_mode.py:141: UserWarning: User provided device_type of 'cuda', but CUDA is not available. Disabling\n",
      "  warnings.warn('User provided device_type of \\'cuda\\', but CUDA is not available. Disabling')\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(Betty,)"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "doc = nlp(\"I visited my friend Betty at her house\")\n",
    "doc.ents"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "7a8798c9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "PRON 4690420944186131903\n",
      "VERB 15806774029700244804\n",
      "PRON 227504873216781231\n",
      "NOUN 16302678419497547123\n",
      "PROPN 3364724584372448665\n",
      "ADP 11667289587015813222\n",
      "PRON 4115755726172261197\n",
      "NOUN 9471806766518506264\n"
     ]
    }
   ],
   "source": [
    "for word in doc:\n",
    "    print(word.pos_, word.lemma)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "98b45a3d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "WordpieceBatch(strings=[['<s>', 'It', 'Ġwent', 'Ġthere', 'Ġunw', 'ill', 'ingly', '.', '</s>']], input_ids=array([[    0,   243,   439,    89, 10963,  1873,  7790,     4,     2]],\n",
       "      dtype=int32), attention_mask=array([[1., 1., 1., 1., 1., 1., 1., 1., 1.]], dtype=float32), lengths=[9], token_type_ids=None)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "doc = nlp(\"It went there unwillingly.\")\n",
    "doc._.trf_data.wordpieces \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "3a621442",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[[ 0.33051595, -0.8022069 , -1.6497264 , ...,  0.6581548 ,\n",
       "          0.9561035 , -0.2500044 ],\n",
       "        [-1.181412  ,  0.20144682,  0.00258117, ..., -0.6660982 ,\n",
       "          0.19771133, -0.77598953],\n",
       "        [-0.7954974 , -0.769232  , -0.39256045, ..., -2.1504755 ,\n",
       "         -0.93583614, -0.13441502],\n",
       "        ...,\n",
       "        [ 0.77691317, -0.75669503, -0.9464762 , ..., -1.2225597 ,\n",
       "          1.2749367 ,  0.34522754],\n",
       "        [ 0.31343666, -0.77399063, -1.7049032 , ...,  0.598163  ,\n",
       "          0.9535197 , -0.24641901],\n",
       "        [ 0.3133781 , -0.7740551 , -1.7045587 , ...,  0.5984289 ,\n",
       "          0.95377105, -0.2467697 ]]], dtype=float32)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "doc._.trf_data.tensors[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "baf4fc60",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1, 9, 768)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "doc._.trf_data.tensors[0].shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9b0dd27c",
   "metadata": {},
   "source": [
    "## Chapter 10"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2673289f",
   "metadata": {},
   "source": [
    "### Extracting city entities"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "39b74c58",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "San Jose \t I would like for it to be in San Jose.\n",
      "San Jose \t I see that at 71 Saint Peter there is a good restaurant which is in San Jose.\n",
      "Bazille \t How would you like Bazille restaurant which is situated in San Jose.\n",
      "San Jose \t How would you like Bazille restaurant which is situated in San Jose.\n",
      "Palo Alto \t Do you have another restaurant matching my needs? For example a restaurant which is economical and is located in Palo Alto.\n",
      "Palo Alto \t I see that 7 restaurants suit to what you requested. Bird Dog seems as a good restaurant and is located in Palo Alto.\n",
      "Palo Alto \t Can you please confirm that you want to book a table for 2 at 11:30 am at the Bird Dog restaurant in Palo Alto for today.\n",
      "milpitas \t Some Punjabi kind of foods in milpitas\n",
      "Milpitas \t I found a good restaurant in Milpitas, The restaurant name is Anjappar Chettinad Restaurant\n",
      "Milpitas \t I found another restaurant for you, Aria Dining & Banquets Fine Indian Cuisine Restaurant at Milpitas\n",
      "Milpitas \t I found 2 Restaurants, Olive garden Italian Restaurant at Milpitas\n",
      "Milpitas \t Please confirm the following details: Booking a table at Olive Garden Italian Restaurant at Milpitas on March 11th reservation time is 1:30 pm, The reservation is for 2 people\n",
      "Berkeley \t Find me Ethiopian cuisine in Berkeley.\n",
      "Berkeley \t Of the 4 Ethiopian restaurants in Berkeley, Addis Restaurant is a good one.\n",
      "Addis Restaurant \t Of the 4 Ethiopian restaurants in Berkeley, Addis Restaurant is a good one.\n",
      "Berkeley \t Please confirm that you want to reserve a table for 2 at the Addis Restaurant in Berkeley for tomorrow at 5:15 pm.\n",
      "Berkeley \t Okay.Please confirm that you want to reserve a table for 2 at the Addis Restaurant in Berkeley for tomorrow at 5:30 pm.\n",
      "Palo Alto \t Some Freshwater fish kind of food in Palo Alto would be perfect.\n",
      "Palo Alto \t I found 4 restaurants. Odori Japanese Cuisine is a nice restaurant in Palo Alto.\n",
      "Palo Alto \t P.f. Chang's is another nice restaurant in Palo Alto.\n",
      "Palo Alto \t Please confirm: Booking a table for 1 at P.f. Chang's at 6:30 pm today in Palo Alto.\n",
      "San Mateo \t Lobster sounds great. Please search in San Mateo.\n",
      "San Mateo \t I have found 9 restaurants. Andes Cafe San Mateo is a great restaurant in San Mateo.\n",
      "San Mateo \t Fuji Sukiyaki is another great choice in San Mateo.\n",
      "San Mateo \t Golden Wok would be a great choice in San Mateo.\n",
      "San Mateo \t Please confirm the following: You want to eat at Golden Wok in San Mateo on March 14th starting at 5:30 pm. The reservation would be for 2 people.\n",
      "San Bruno \t I want to eat in San Bruno\n",
      "San Bruno \t I have 3 restaurants, Cabalen is a nice one in San Bruno.\n",
      "San Bruno \t What about Isla Restaurant in San Bruno?\n",
      "San Mateo Avenue \t It is moderate pricing and they are at 448 San Mateo Avenue\n",
      "San Bruno \t Please confirm: Booking a table for 2 at Isla Restaurant in San Bruno. It is on March 12th at 6 pm.\n",
      "San Jose \t The city is San Jose.\n",
      "San Jose \t I have one in San Jose called Ariake.\n",
      "Ariake \t I have one in San Jose called Ariake.\n",
      "San Jose \t Please confirm Ariake in San Jose for 6 pm on March 10th for 2.\n",
      "San Jose \t My family and I are in San Jose and the perfect place would have pizza and pasta. We don't have a price preference.\n",
      "San Jose \t San Jose has a restaurant called Al Castello Ristorante that is nice.\n",
      "San Jose \t There is also a nice restaurant called Angelo's Italian Restaurant in San Jose.\n",
      "San Jose \t There is another nice restaurant called Antonella's Ristorante in San Jose.\n",
      "San Jose \t I will be booking a table at Antonella's Ristorante in San Jose. The reservation will be set at 12:30 pm for today. And it is for 2 people is this correct.\n",
      "Pleasanton \t Can you find me something economical to eat in Pleasanton.\n",
      "Pleasanton \t Found 5 restaurants for you out of which De La Torre's Trattoria seems a great from Pleasanton.\n",
      "n Pleasanton \t Forno Vecchio also seems nice and its located n Pleasanton as well.\n",
      "Pleasanton \t Another good restaurant in Pleasanton in Gay 90's Pizza Co.\n",
      "Pleasanton \t Please confirm that you want to make a reservation at Gay 90's Pizza Co. which is located in Pleasanton for 2 people. The reservation is for today at 1:30 pm.\n",
      "Fairfield \t I want some breakfast in Fairfield\n",
      "fairfield \t I recommend Mimi's cafe in fairfield\n",
      "Fairfield \t Please confirm your booking at mimi's cafe in Fairfield for 1 people today at 4:45 pm\n",
      "Fairfield \t I'm starving! Can you help me find a restaurant that serves Breakfast in Fairfield?\n",
      "Fairfield \t There is 1 restaurant that meets your needs. Mimi's Cafe in Fairfield serves breakfast.\n",
      "Fairfield \t You'd like a table for 2 today at Mimi's Cafe in Fairfield at 5:30 pm. Is this correct?\n",
      "Cupertino \t Cupertino, please.\n",
      "Cupertino \t I've found 2 Malaysian restaurants in Cupertino. Merlion Restaurant & Bar is one.\n",
      "Cupertino \t Rasa Sayang is also is Cupertino.\n",
      "Cupertino \t So you'd like a reservation for 2 at 5:15 pm today, at Rasa Sayang in Cupertino?\n",
      "SF \t I would prefer a restaurant which serves up Cambodian food in SF.\n",
      "Anh Hong \t I have managed to find a nice restaurant called Anh Hong which is located in San Francisco.\n",
      "San Francisco \t I have managed to find a nice restaurant called Anh Hong which is located in San Francisco.\n",
      "San Francisco \t Yes there is, Aux Delices is also a nice restaurant located in San Francisco.\n",
      "San Francisco \t Please confirm your reservation for 2 people at Aux Delices in San Francisco at 7:15 pm today.\n",
      "Fremont City \t Asian kind of food will be good for me and please search it in Fremont City.\n",
      "Fremont city \t Found 3 results, Asian pearl Seafood Restaurant is the best one in Fremont city, hope you like it.\n",
      "Fremont \t Please confirm to book a table for 2 persons in Asian Pearl Seafood REstaurant in Fremont at 1:30 pm on today.\n",
      "Albany \t I want to find a restaurant in Albany.\n",
      "Albany \t I have 2 options. Cafe Eugene in Albany is a good choice.\n",
      "Albany \t Juanita & Maude in Albany is great.\n",
      "Albany \t Making a dinner reservation for 2 people at Juanita & Maude on March 13th at 5 pm in Albany. Is that Correct?\n",
      "San Francisco \t Sure, I found 8 Immortals Restaurant in San Francisco\n",
      "San Francisco \t Okay, got it. I will book a table at 8 Immortals Restaurant in San Francisco for 5:30 pm today, for 6 people.\n",
      "San Francisco \t Certainly, please confirm: Reservation at 8 Immortals Restaurant in San Francisco is for 12 pm today for 4 people.\n",
      "Oakland \t Oh, I'm in Oakland. Please find an American restaurant.\n",
      "Oakland \t I'd recommend Chop Bar in Oakland.\n",
      "Oakland \t Sure, please confirm that you'd like to book a table for 2 at Chop Bar in Oakland from 6 pm today.\n",
      "Oakland \t In Oakland.\n",
      "Oakland \t I found 8 places. A16 is a nice restaurant in Oakland.\n",
      "Oakland \t Just to confirm, you want a table today at A16 in Oakland for 2 people at 5:30 pm?\n",
      "Concord \t Asian Fusion in Concord.\n",
      "Concord \t Ozora Sushi is a restaurant in Concord.\n",
      "Concord \t Ozora Sushi in Concord today for 2 people at 5 pm.\n",
      "Rohnert Park \t Are there any Mexican places in Rohnert Park?\n",
      "Rohnert Park \t Chili's Grill & Bar is in Rohnert Park.\n",
      "Kyoto \t 1 restaurant named Kyoto fits that description. It's in Rohnert Park as well.\n",
      "Rohnert Park \t 1 restaurant named Kyoto fits that description. It's in Rohnert Park as well.\n",
      "Kyoto \t Just to confirm: I am booking a table for 3 today at 11:15 am at the Rohnert Park location of Kyoto.\n",
      "San Fran \t I would like something that I can pick-up in San Fran at any price range.\n",
      "San Francisco \t There is a restaurant called Golden Gate Indian Cuisine & Pizza that is located in San Francisco.\n",
      "Hunan Empire Restaurant \t I was able to locate 1 other nice restaurant called Hunan Empire Restaurant located in San Francisco.\n",
      "San Francisco \t I was able to locate 1 other nice restaurant called Hunan Empire Restaurant located in San Francisco.\n",
      "San Francisco \t Great, I have a dining reservation for 2 people, located at the Hunan Empire Restaurant in San Francisco, on March 13th at 12 pm. Does that sound correct?\n",
      "Redwood City \t I prefer Redwood City.\n",
      "Redwood City \t Redwood City has 1 Spanish restaurant, Estampas Peruanas Restaurant.\n",
      "Millbrae \t I am looking for restaurants located in Millbrae.\n",
      "Millbrae \t I found 1 result, Ipoh Garden Malaysian Cuisine located in Millbrae.\n",
      "Millbrae \t Okay. I have an 11 AM reservation for 2 people today, at Ipoh Garden Malaysian Cuisine in Millbrae. Is that correct?\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Calistoga \t Please look for a place in Calistoga.\n",
      "Calistoga \t There's a nice place called Calistoga Thai Kitchen in Calistoga.\n",
      "Lovina \t I've found 1 restaurant that could work. It's Lovina in Calistoga.\n",
      "Calistoga \t I've found 1 restaurant that could work. It's Lovina in Calistoga.\n",
      "Lovina \t Okay, so it's a reservation at Lovina in Calistoga at 12:30 pm today for 2 people?\n",
      "Calistoga \t Okay, so it's a reservation at Lovina in Calistoga at 12:30 pm today for 2 people?\n",
      "Burlingame \t I want to eat in Burlingame.\n",
      "Burlingame \t There are 2 possible restaurants. One of them is Grand Harbor located in Burlingame.\n",
      "Burlingame \t Please confirm: Booking a table for 2 people at 12:45 pm for today at the Grand Harbor in Burlingame.\n",
      "Rohnert Park \t I want something in Rohnert Park\n",
      "Rohnert Park \t There is a lovely restaurant in Rohnert Park called Hana\n",
      "Pleasant Hill \t There are 2 options available in Pleasant Hill. I would like to recommend Matsu Sushi Japanese Restaurant.\n",
      "Pleasant Hill \t I would like to confirm the following information: a table for 2 today at 8 pm at the Matsu Sushi Japanese Restaurant in Pleasant Hill.\n",
      "San Ramon \t I'd like some soup, and I'm looking in San Ramon.\n",
      "San Ramon \t Noodles And Company is a nice resturant in San Ramon.\n",
      "San Ramon \t I have booked a table at Noodles And Company in San Ramon at 7:30 pm today for 2 people. Is that correct?\n",
      "Cupertino \t Search in Cupertino for Dumplings.\n",
      "Cupertino \t Cupertino has a dumpling restaurant called A Plus Cafe.\n",
      "Santa Clara \t Can you look for another place to eat in Santa Clara?\n",
      "Santa Clara \t There's a nice place in Santa Clara called Chef Ming's.\n",
      "Santa Clara \t How about China Stix in Santa Clara?\n",
      "Santa Clara \t Can you confirm the details for me? You're booking a table for 2 next Thursday at 7:30 pm at China Stix in Santa Clara.\n",
      "San Mateo \t I would love to taste some Seafood in San Mateo. Please find some good hotels.\n",
      "San Mateo \t Please give me a moment. There are 9 restaurants with good rating and the top among them is Andes Cafe San Mateo located in San Mateo. Would you prefer to have your lunch here?\n",
      "San Mateo \t Please give me a moment. There are 9 restaurants with good rating and the top among them is Andes Cafe San Mateo located in San Mateo. Would you prefer to have your lunch here?\n",
      "San Mateo \t Okay! Confirm me the details that you wish to book a table at Andes Cafe San Mateo in San Mateo at 1:15 pm by today and the reservation would be for 2 persons, right?\n",
      "Dublin \t I'm looking for somewhere in Dublin to have a meal and a drink.\n",
      "Dublin \t I found 2 restaurants. One is the Black Angus Steakhouse in Dublin.\n",
      "Dublin \t Hot Crab is also located in Dublin.\n",
      "Dublin \t Okay. Please confirm that you'd like a booking for 3 at Hot Crab in Dublin today at 5 pm.\n",
      "Napa \t I need Pizza kind of food and i wish to eat in city Napa.\n",
      "Napa \t I found 1 Restaurant named as Bistro Don Giovanni which is located in Napa\n",
      "Napa \t ok here are the details please check or review it . Table booked at Bistro Don Giovanni which is in city Napa. Reservation time is 11:30 am for 2 people and the date is today.\n",
      "Milpitas \t Milpitas.\n",
      "Milpitas \t There is a Cambodian restaurant in Milpitas called Anh Hong.\n",
      "Anh Hong \t There is a Cambodian restaurant in Milpitas called Anh Hong.\n",
      "Anh Hong \t Sure. A reservation at Anh Hong in Milpitas for 4 at 6:30 pm today. Is that correct?\n",
      "Milpitas \t Sure. A reservation at Anh Hong in Milpitas for 4 at 6:30 pm today. Is that correct?\n",
      "Anh Hong \t Perfect. What are the prices at Anh Hong like?\n",
      "Martinez \t I'd like to go to a restaurant in Martinez that serves Thai food.\n",
      "Martinez \t Okay, I found 1 nice restaurant in Martinez called Lemongrass Bistro.\n",
      "Martinez \t Alright, please confirm that you'd like me to book a table for 2 at Lemongrass Bistro in Martinez for today at 7 pm.\n",
      "Pleasanton \t Any place in Pleasanton which serves Ramen kind of food and also alcohol will be ideal.\n",
      "Ramen \t Any place in Pleasanton which serves Ramen kind of food and also alcohol will be ideal.\n",
      "Pleasanton \t There is a nice restaurant called Meiko Sushi in Pleasanton.\n",
      "Pleasanton \t So, a table reservation at Meiko Sushi in Pleasanton will be for 12 pm on the day after tomorrow for 6 people. Please confirm the aforementioned details.\n",
      "Pleasanton \t I would like to it in a place in Pleasanton.\n",
      "Pleasanton \t I have for you 3 option .One is a nice restaurant located in Pleasanton named Andy & Yu's.\n",
      "Pleasanton \t Book a table at Andy & Yu's in Pleasanton at 6:45 pm today for 2 person is correct?\n",
      "Palo Alto \t Palo Alto.\n",
      "Palo Alto \t I found Pampas in Palo Alto.\n",
      "Palo Alto \t Please confirm a table at Pampas in Palo Alto at 5:15 pm for 2 people today.\n",
      "Milpitas \t I want to go to Milpitas. Spicy Indian food sounds perfect.\n",
      "Milpitas \t I found 4 restaurants- how about Anjappar Chettinad Restaurant in Milpitas?\n",
      "Milpitas \t So I'll book a table for 4 at Anjappar Chettinad Restaurant in Milpitas for 1 pm today. Is that okay?\n",
      "Milpitas \t I would like to eat in Milpitas and the restaurant should be in the moderate price range.\n",
      "Milpitas \t There are 2 restaurants in Milpitas I have found. The Olive Garden Italian Restaurant meets your criteria.\n",
      "Milpitas \t The other restaurant in Milpitas is Red Chillies The Malabar Cuisine.\n",
      "Milpitas \t The reservation is for today at Red Chillies The Malabar Cuisine at 11 am in Milpitas. The reservation is for 2 people. Is this correct?\n",
      "Milpitas \t The reservation is in Milpitas on March 14th. Is this correct?\n",
      "Dublin \t Look for a place in Dublin.\n",
      "Dublin \t There are 7 restaurants that could work. There's a nice one called China Paradise Restaurant in Dublin.\n",
      "Dublin \t Your reservation will be for 1 person next Thursday at 1 pm at China Paradise Restaurant in Dublin. Is that correct?\n",
      "Petaluma \t I want something with a moderate priced in Petaluma. It also needs to serve alcohol.\n",
      "Petaluma \t I found 1 restaurant, Shree Indian Cuisine. It's a nice restaurant in Petaluma.\n",
      "Petaluma \t Okay, making a reservation for 2 people Shree Indian Cuisine in Petaluma at 11 am today. Is that correct?\n",
      "Petaluma \t Okay, making the reservation for 1 person at Shree Indian Cuisine in Petaluma.\n",
      "San Francisco \t I have to have something in San Francisco.\n",
      "San Francisco \t I have a good option, try 54 Mint Ristorante Italiano in San Francisco.\n",
      "San Francisco \t Let me clarify your plans: You want a table at 54 Mint Ristorante Italiano in San Francisco at 5:15 pm. today for 2?\n",
      "Foster City \t Let's go with Foster City please.\n",
      "Foster City \t One moment. I found a great restaurant called Pastries N Chaat in Foster City.\n",
      "Foster City \t How about, Tabla Indian Restaurant in Foster City?\n",
      "Foster City \t One moment. Please confirm a reservation for 2 people, at Tabla Indian Restaurant in Foster City. At 1:45 pm today. Does this sound right?\n",
      "San Mateo \t Search in San Mateo.\n",
      "San Mateo \t There are 5 restaurants, Gau Poang is a nice restaurant located in San Mateo.\n",
      "SF \t Can you suggest other restaurants in SF?\n",
      "San Francisco \t Alice's restaurant is located in San Francisco.\n",
      "San Francisco \t Bamboo Restaurant is located in San Francisco.\n",
      "San Francisco \t Please confirm you are booking a table for 2 at Bamboo Restaurant, located in San Francisco today at 12 pm.\n",
      "California \t Food kind of California will be perfect in SF.\n",
      "SF \t Food kind of California will be perfect in SF.\n",
      "San Francisco \t There are 10 restaurants, Al's Place is one of the good restaurant in San Francisco.\n",
      "San Francisco \t Alta Msp is one of the good restaurant in San Francisco.\n",
      "San Francisco \t Aq is one of the good restaurant in San Francisco.\n",
      "San Francisco \t Please confirm the details: You want me to make a reservation at Aq in San Francisco for today at 7:30 pm for 2 people.\n",
      "Foster City \t Some Izakaya kind of food good in Foster City\n",
      "Japanese Restaurant \t I have found out 2 restaurant for you and one of them is Tokai Japanese Restaurant located in Foster city is a nice one.\n",
      "Foster \t I have found out 2 restaurant for you and one of them is Tokai Japanese Restaurant located in Foster city is a nice one.\n",
      "Foster City \t The restaurant Tokyo Village Grill & Sushi located in Foster City is a nice one.\n",
      "Foster City \t please give the confirmation details:Booking the table at Tokyo Village Grill & Sushi at Foster City and the reservation is for 2 people at 6:30 pm on next monday\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "San Francisco \t I like Pizza and Pasta kind of food and the place should be in San Francisco.\n",
      "San Francisco \t There are 10 restaurants. One among is 54 Mint Ristorante Italiano which is a nice restaurant located in San Francisco\n",
      "San Francisco \t Ok Please check and verify the following details so that i will confirm. Table has been booked at 54 Mint Ristorante Italiano in the city San Francisco and the reservation time is at 6 pm for 2 people and the date is today.\n",
      "Pleasanton \t Some Punjabi food would be great. Please look in Pleasanton.\n",
      "Pleasanton \t I found 7 restaurants in Pleasanton. The first one is Aappakadai.\n",
      "Pleasanton \t Please confirm the following details: Booking a table for 2 at Aappakadai in Pleasanton for today at 12:15 pm.\n",
      "Sunnyvale \t I'd like pizza and somewhere in Sunnyvale.\n",
      "Sunnyvale \t I found 3 places. Gumba's is a good place in Sunnyvale.\n",
      "Sunnyvale \t Just to confirm, you want a table at Gumba's in Sunnyvale for 2 people today at 7:15 pm?\n",
      "San Fran \t Look in San Fran.\n",
      "San Francisco \t There are 10 restaurants that fit what you're looking for. 1760 is a nice restaurant that's in San Francisco.\n",
      "San Francisco \t Okay, I'll you want a reservation at 1760 in San Francisco for 2 people, today at 4:45 pm, is that correct?\n",
      "San Jose \t Something Southern in the San Jose area would be great.\n",
      "San Jose \t There's a nice restaurant by the name of 71 Saint Peter in San Jose.\n",
      "San Jose \t Hukilau is another nice restaurant in San Jose that meets that criteria.\n",
      "San Jose \t Another nice restaurant in San Jose is Rosie Mccann's Irish Pub & Restaurant.\n",
      "San Jose \t To confirm: You'd like me to make a reservation for today a 11:45 am for 2 guests at Rosie Mccann's Irish Pub & Restaurant in San Jose, correct?\n",
      "Oakland \t I prefer it in Oakland. I am craving Pizza.\n",
      "Oakland \t Your search found 4 restaurants. One is Belotti Ristorante E Bottega which is located in Oakland.\n",
      "Oakland \t Contrasto is a nice restaurant and located in Oakland.\n",
      "Oakland \t Please say yes to confirm the following. Booking a table at Contrasto in Oakland with a reservation at 11:30 am today. The reservation is a party of 2.\n",
      "Oakland \t Please say yes to confirm the following. The city is Oakland, and the reservation is for a party of 4.\n",
      "San Francisco \t Bistro Aix is a good restaurant in San Francisco\n",
      "San Francisco \t Confirm Bistro Aix and the city is San Francisco. Reservation os 1:15 pm and for 2 people today\n",
      "Dublin \t Please look in Dublin, a burger is perfect to have now.\n",
      "Dublin \t I have found 1 restaurant that you might like, maybe you like McDonald's in Dublin?\n",
      "Dublin \t Please do confirm the following details: book seats at McDonald's in Dublin at 7:15 pm for 2 people today.\n",
      "Dublin \t Please do confirm the following details: book seats at McDonald's in Dublin at 7 pm.\n",
      "Dublin \t Please do confirm the following details: book seats at McDonald's in Dublin for 3 people on March 14th.\n",
      "Milpitas \t I would like to stay around Milpitas.\n",
      "Milpitas \t There is a great restaurant called Baja Cactus in Milpitas.\n",
      "Santa Clara \t That's great, are there any other restaurants, maybe in Santa Clara?\n",
      "Santa Clara \t Santa Clara has a nice restaurant named El Amigo Burrito.\n",
      "Santa Clara \t Santa Clara also has a great restaurant called La Paloma.\n",
      "Santa Clara \t Please confirm your reservation. I have booked a table for 2 at La Paloma in Santa Clara at 12:30 pm today. Is that correct?\n",
      "San Francisco \t Do you know of any places to eat in San Francisco?\n",
      "San Francisco \t There's a nice restaurant called 1760 in San Francisco.\n",
      "San Francisco \t There are 10 restaurants that match. There's a nice restaurant called Academy Bar and Kitchen that's in San Francisco.\n",
      "San Francisco \t So you'd like a 12:45 pm reservation today at Academy Bar and Kitchen in San Francisco for 2 people, is that right?\n",
      "Oakland \t In Oakland.\n",
      "Oakland \t I found 7 restaurants matching this query. One is Alamar Kitchen & Bar in Oakland.\n",
      "Oakland \t There's Judoku Sushi in Oakland.\n",
      "Oakland \t I found 4 restaurants to match. One is Lake Chalet in Oakland.\n",
      "Oakland \t Please confirm: Table for 2 at Lake Chalet, 10:30 am today in Oakland.\n",
      "Oakland \t Please confirm: Table for 1 at Lake Chalet, 12:30 pm today in Oakland.\n",
      "Alameda \t Find me a sushi place in Alameda.\n",
      "Alameda \t I have found a restaurant called Angel Fish in Alameda.\n",
      "Alameda \t Please confirm the following: You want a table at Angel Fish in Alameda at 1:15 pm today for 2 people.\n",
      "Millbrae \t I'm in the mood for some Chinese food tonight, preferably in Millbrae.\n",
      "Millbrae \t I managed to find 1 restaurant in Millbrae, it's called Tai Wu Restaurant. I think you'd like it.\n",
      "Millbrae \t So you'd like a table for today at Tai Wu Restaurant for 2 people at 7:30 pm in Millbrae. Is that correct?\n",
      "Berkeley \t I want some Noodles, and search in Berkeley.\n",
      "Berkeley \t Upon searching I have found 6 restaurants. There's one called Crave Asia which is pretty nice and situated in Berkeley.\n",
      "Berkeley \t Okay, please confirm the details one more time: Booking a table for 2 people at Crave Asia in Berkeley, the reservation should be for 11:30 AM and the date is today, correct?\n",
      "Petaluma \t I was thinking in Petaluma. Preferably a Pizza place.\n",
      "Petaluma \t I've found 4 in Petaluma, Brixx Pizzeria is a nice restaurant.\n",
      "Petaluma \t Cucina Paradiso is also a nice one in Petaluma.\n",
      "Petaluma \t Risibisi is another nice restaurant in Petaluma.\n",
      "Petaluma \t Please confirm that you are booking a table at risibisi in Petaluma. Your reservation will be at 1:30 pm. Set for today for 2 people.\n",
      "San Fran \t I'm looking for brunch in San Fran.\n",
      "Fillmore \t There is a nice restaurant named 1300 On Fillmore located in San Francisco.\n",
      "San Francisco \t There is a nice restaurant named 1300 On Fillmore located in San Francisco.\n",
      "San Francisco \t There is another nice restaurant named the Absinthe Brasserie & Bar in San Francisco.\n",
      "San Francisco \t Aperto is another nice restaurant in San Francisco.\n",
      "Aperto \t Please confirm you want me to reserve a table for 2 people at 11 am at Aperto in San Francisco today.\n",
      "San Francisco \t Please confirm you want me to reserve a table for 2 people at 11 am at Aperto in San Francisco today.\n",
      "Pleasanton \t I need to find somewhere to get a good meal in Pleasanton, can you help?\n",
      "Osaka \t I want Japanese food, my girlfriend is from Osaka, I'm trying to impress her.\n",
      "Pleasanton \t Okay, I found 4 good Japanese places. How about Meiko Sushi in Pleasanton?\n",
      "Pleasanton \t I found 2 affordable places in Pleasanton. How about P.f. Chang's?\n",
      "Pleasanton \t Okay, please confirm that you'd like to book a table for 2 at P.F. Chang's in Pleasanton for 7:45 pm today.\n",
      "Petaluma \t Do you know a place to eat in Petaluma?\n",
      "Petaluma \t In Petaluma there are 4 restaurants. Brixx Pizzeria is a nice one.\n",
      "Petaluma \t You want to make a reservation at 5 pm today for 6 people at Brixx Pizzeria in Petaluma?\n",
      "Pleasanton \t I would prefer to eat at Pleasanton.\n",
      "Pleasanton \t There is a nice restaurant called Meiko Sushi located in Pleasanton.\n",
      "Pleasanton \t Please confirm your reservation for 2 at Meiko Sushi in Pleasanton at 12:30 pm today.\n",
      "Pleasanton \t Sure! Please confirm your reservation for 2 at Meiko Sushi in Pleasanton at 7:15 pm today.\n",
      "Pleasanton \t My apologies, I was not able to make the reservation. Alternatively, would you like a for 2 at Meiko Sushi in Pleasanton at 7:30 pm today instead? They are located at 4301 Valley Avenue B.\n",
      "San Francisco \t I've found 10 possible options - Amber in San Francisco has 4.2 stars.\n",
      "San Francisco \t To review: You'd like to make an 11 am reservation at Amber restaurant in San Francisco for 2 people, today.\n",
      "San Jose \t I have a friend visiting from out of town and need to find somewhere good for us to eat. Can you look up restaurants in San Jose?\n",
      "San Jose \t I found 10 possible options in San Jose. How about Ariake?\n",
      "Ariake \t Sure. Just to confirm, you'd like to make a reservation for 2 people at Ariake in San Jose on March 10th at 6:45 pm. Is that right?\n",
      "San Jose \t Sure. Just to confirm, you'd like to make a reservation for 2 people at Ariake in San Jose on March 10th at 6:45 pm. Is that right?\n",
      "Ariake \t Sure. Just to confirm, you'd like to make a reservation for 2 people at Ariake in San Jose on March 10th at 1 pm. Is that right?\n",
      "San Jose \t Sure. Just to confirm, you'd like to make a reservation for 2 people at Ariake in San Jose on March 10th at 1 pm. Is that right?\n",
      "Ariake \t Ok, just to confirm, I'm making the reservation at Ariake for next Tuesday.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "San Francisco \t It will be fine at San Francisco and pizza and pasta will cherish me.\n",
      "San Francisco \t Then 54 Mint Ristorante Italiano in San Francisco will be the right one.\n",
      "San Francisco \t Please confirm, Today at 5:30 pm in San Francisco, table 54 Mint Ristorante Italiano for 2 people.\n",
      "San Francisco \t We are meeting in San Francisco.\n",
      "San Francisco \t There is a nice restaurant in San Francisco called Ala Romana.\n",
      "San Francisco \t Che Fico is another nice restaurant in San Francisco.\n",
      "Cala \t There are 3 Mexican restaurants in the area. Cala is a nice one in San Francisco.\n",
      "San Francisco \t There are 3 Mexican restaurants in the area. Cala is a nice one in San Francisco.\n",
      "Cala \t I want to make a reservation for Cala.\n",
      "Cala \t To confirm, the reservation is at Cala in San Francisco at 1 pm next Thursday for 2 people.\n",
      "San Francisco \t To confirm, the reservation is at Cala in San Francisco at 1 pm next Thursday for 2 people.\n",
      "Cala \t They do not have live music. I was unable to make the reservation at Cala, shall I try to reserve next Thursday at 12 pm for 2?\n",
      "Milpitas \t I want to go out to eat in Milpitas.\n",
      "Milpitas \t I found 2 possibilities. There is an Olive Garden Italian restaurant in Milpitas.\n",
      "Milpitas \t Please confirm the following: You want a table for 3 at Olive Garden Italian Restaurant in Milpitas today at 1 pm?\n",
      "Milpitas \t Please confirm the following details: You want a table for 4 at Olive Garden Italian Restaurant in Milpitas today at 1 pm?\n",
      "San Jose \t If I got any place in San Jose then it will be nice for me.\n",
      "San Jose \t found 1 result, a restaurant in San Jose and its name is Mcdonald's.\n",
      "San Jose \t Please confirm to book a table for 2 persons at Mcdonald's in San Jose at 11:30 am on day after tomorrow.\n",
      "San Jose \t Please confirm to book the table for 1 person in San Jose.\n",
      "Santa Clara \t I'm looking to have Korean Barbeque. I'm looking in Santa Clara.\n",
      "Santa Clara \t There is 1 restaurant that matches that. There's a nice restaurant called Restaurant Chungdam in Santa Clara.\n",
      "Santa Clara \t Great. Please verify that the following is correct: one table to Restaurant Chungdam in Santa Clara at 8 pm today for 2 people.\n",
      "Santa Clara \t Great. Please verify that the following is correct: one table to Restaurant Chungdam in Santa Clara at 7 pm today for 2 people.\n",
      "San Francisco \t I've found 4 potential restaurants for you. How about Izakaya Sushi Ran in San Francisco?\n",
      "San Francisco \t Okay, please confirm: Table for 2 at Izakaya Sushi Ran in San Francisco today at 5:45 pm.\n",
      "Berkeley \t I'd like to eat somewhere in Berkeley.\n",
      "Berkeley \t The B&b Kitchen & Wine Bar is a great restaurant in Berkeley and a local favorite.\n",
      "Berkeley \t I have booked a table for 2 at B&b Kitchen & Wine Bar for today at 1:30 pm in Berkeley. Is this correct?\n",
      "Berkeley \t I have booked a table for 2 at B&b Kitchen & Wine Bar for today at 7 pm in Berkeley. Is this correct?\n",
      "San Mateo \t I want to eat in San Mateo.\n",
      "San Mateo \t There's a nice restaurant in San Mateo called Celia's.\n",
      "San Mateo \t Please confirm: Making a reservation for 2 people at Celia's in San Mateo for today at 11:30 am.\n",
      "San Francisco \t I found 10 restaurants. One of them is Baker Street Bistro in San Francisco.\n",
      "San Francisco \t Blue Mermaid in San Francisco.\n",
      "San Francisco \t Please confirm Blue Mermaid in San Francisco at 5:30 pm for 2 people today.\n",
      "Oakland \t I'd like to get some Middle eastern food in Oakland.\n",
      "Oakland \t There's a good restaurant in Oakland called Dyafa.\n",
      "Oakland \t I found 1 that might work for you. It is called Marhaba Indian & Pakistani Halal Cuisine and it is in Oakland.\n",
      "Oakland \t Okay, so you would like to reserve at table at Marhaba Indian & Pakistani Halal Cuisine in Oakland, for 2 people today at 1:30 pm. Is that correct?\n",
      "Calistoga \t Ya, I like to eat at some nice place in Calistoga.\n",
      "Calistoga \t I got details about 1, Soo Yuan Restaurant is the best one which is located in Calistoga as per your wish.\n",
      "Calistoga \t Please confirm: Booking for a table for 4 persons at Soo Yuan Restaurant in Calistoga at 1:30 pm today.\n",
      "Berkeley \t I will be in Berkeley and am in the mood for some Breakfast type of food.\n",
      "Berkeley \t I found 1 match called the Venus Restaurant in Berkeley.\n",
      "Berkeley \t You would like to book a table for 1 at 12:15 pm today at the Venus Restaurant in Berkeley. Is that correct?\n",
      "Sushi \t I would like a reasonably priced restaurant that serves Sushi in San Rafael.\n",
      "San Rafael \t I would like a reasonably priced restaurant that serves Sushi in San Rafael.\n",
      "San Rafael \t There is a great restaurant called Sushi To Dai For in San Rafael.\n",
      "San Rafael \t There is a great restaurant that serves Sichuan cuisine in San Rafael called Yet Wah Restaurant.\n",
      "San Rafael \t Excellent. Can you make a reservation at Yet Wah Restaurant in San Rafael for 7 pm next Monday for 2 people?\n",
      "Cotati \t Please book me a restaurant in Cotati.\n",
      "Cotati \t There is a restaurant called Mi Pueblo in Cotati that has good tacos.\n",
      "Cotati \t Can you find me another restaurant in Cotati with live music?\n",
      "Cotati \t There's 1 restaurant in Cotati that fits your criteria. It is called Mi Ranchito.\n",
      "Mi Ranchito \t There's 1 restaurant in Cotati that fits your criteria. It is called Mi Ranchito.\n",
      "Cotati \t Please verify the following details. You would like to reserve a table for 2 today at 6:30 pm at Mi Ranchito in Cotati.\n",
      "San Francisco \t Show me places to eat in San Francisco. I am looking for something afforadable priced.\n",
      "San Francisco \t I found 10 restaurants that serve pizza and pasta. 54 Mint Ristorante Italiano is a nice restaurant located in San Francisco.\n",
      "San Francisco \t Please confirm the following details: You are booking a table at 54 Mint Ristorante Italiano, the city is San Francisco, the reservation is at 5:30 pm, the date is the day after tomorrow, and the reservation is for 4 people.\n",
      "SF \t Somewhere in SF, and I am really craving Thai food at the moment!\n",
      "San Francisco \t I found a bunch of restaurants, there's actually 10 that you might like in San Francisco, one of them being Baan Thai House & Wine Bar\n",
      "San Francisco \t Confirming your 6 pm booking today for 2 people at Baan Thai House & Wine Bar in San Francisco.\n",
      "San Bruno \t San Bruno, I am looking for low priced, Taiwanese food.\n",
      "San Bruno \t China Garden fits your preferences in San Bruno.\n",
      "San Mateo Avenue \t 586 San Mateo Avenue\n",
      "San Bruno \t Dumpling Era in San Bruno serves alcohol.\n",
      "San Bruno \t I found 1 restaurant: Ming's in San Bruno.\n",
      "San Bruno \t Please confirm: I am booking a table at Ming's, in San Bruno. The reservation is for 6 people, today at 12:30 pm.\n",
      "San Bruno \t Please confirm: I am booking a table at Ming's in San Bruno, the reservation is at 11 am today for 6 people.\n",
      "San Carlos \t Please locate one in San Carlos.\n",
      "San Carlos \t I found 2 Greek restaurants. Agora is a nice one located in San Carlos.\n",
      "San Carlos \t I located 2 restaurants. Asya Restaurant is nice and located on San Carlos.\n",
      "San Carlos \t I found 1, Thai Time is nice and located in San Carlos.\n",
      "San Carlos \t Please confirm the following: Booking a table for 2 today at Thai Time in the city of San Carlos for 12:45 pm.\n",
      "Petaluma \t I am looking for a restaurant in Petaluma.\n",
      "Petaluma \t How about Flowering Tea House in Petaluma?\n",
      "Petaluma \t Please confirm your reservation details: Reservation for 2 next Friday, at 5:15 PM, in Petaluma at the Flowering Tea House.\n",
      "San Francisco \t I would like a restaurant that serves alcohol in San Francisco. I'd like to eat in a Burgers restaurant please.\n",
      "San Francisco \t There were 8 matches to your inquiry. 5a5 Steak Lounge is a nice place that's in San Francisco.\n",
      "San Francisco \t Alexander's Steakhouse is also a nice place in San Francisco.\n",
      "San Francisco \t There's another restaurant in San Francisco that's called Bourbon Steak Restaurant.\n",
      "San Francisco \t Can you confirm the reservation details please? I am booking a table at Bourbon Steak Restaurant in San Francisco. The reservation is for 1:30 pm for today and is for 2 people.\n",
      "Pleasanton \t Search in Pleasanton.\n",
      "Pleasanton \t There are 2 restaurants I've found. There's a nice restaurant called Baci Bistro And Bar in Pleasanton that might interest you.\n",
      "Pleasanton \t Okay, I'll make you a reservation for 2 at Baci Bistro And Bar in Pleasanton at 1:45 pm today, is that correct?\n",
      "San Francisco \t I'm craving some Sushi in San Francisco if possible.\n",
      "San Francisco \t There a good restaurant in San Francisco that sells sushi it is called 2g Japanese Brasserie.\n",
      "San Francisco \t Please confirm: Making a reservation for 1 person today at 12 pm at 2g Japanese Brasserie in San Francisco.\n",
      "San Francisco \t Please confirm: Making a reservation for 3 people today at 12 pm at 2g Japanese Brasserie in San Francisco.\n",
      "San Francisco \t Making a reservation for 3 people today at 12:30 pm at 2g Japanese Brasserie in San Francisco.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Antioch \t There's a place in Antioch call Okawa Japanese Restaurant.\n",
      "Antioch \t Just to be correct, you're booking a table at Okawa Japanese Restaurant in Antioch at 6 pm. That will be today and that will be for 1.\n",
      "Vallejo \t Find a place in Vallejo with live music.\n",
      "Vallejo \t How about Mcdonald's in Vallejo?\n",
      "Vallejo \t Mcdonald's in Vallejo for 2 at 1 pm today. Correct?\n",
      "SF \t Is there any other restaurant you recommend, preferably in SF?\n",
      "Amelie San Francisco \t I discovered 10 restaurants. There is a wonderful diner called Amelie San Francisco located in San Francisco.\n",
      "San Francisco \t I discovered 10 restaurants. There is a wonderful diner called Amelie San Francisco located in San Francisco.\n",
      "Amelie San Francisco \t Please verify the following information: Reserve a table for 2 people at Amelie San Francisco for 10:30 am today in San Francisco.\n",
      "San Francisco \t Please verify the following information: Reserve a table for 2 people at Amelie San Francisco for 10:30 am today in San Francisco.\n",
      "Amelie San Francisco \t Please verify the following information: reserving a table for 2 people at Amelie San Francisco at 11 am today in San Francisco.\n",
      "San Francisco \t Please verify the following information: reserving a table for 2 people at Amelie San Francisco at 11 am today in San Francisco.\n",
      "San Francisco \t I am in San Francisco.\n",
      "San Francisco \t I have 1 restaurant you may like. Red Tavern is a nice place located in San Francisco.\n",
      "San Francisco \t Please confirm these details: I'm reserving a table for 2 people at Red Tavern in San Francisco, for next Thursday at 6:15 pm.\n",
      "San francisco \t San francisco is okay\n",
      "San francisco \t There is a nice returant called 1760 in San francisco\n",
      "San francisco \t booking table at 1760, city San francisco, time 4:45 pm, date today, reservation for 2 people. Confirm the details above\n",
      "Santa Rosa \t I would like to find a place to eat in Santa Rosa.\n",
      "Santa Rosa \t I have found 4 options. Hikuni is a great choice in Santa Rosa.\n",
      "Santa Rosa \t Osake is a great choice in Santa Rosa.\n",
      "Santa Rosa \t Please confirm the following: You want a table for 2 people at Osake in Santa Rosa at 10:45 am next Monday.\n",
      "San Francisco \t I found 10 restaurants in San Francisco including 8 Immortals Restaurant.\n",
      "San Francisco \t There is a nice restaurant in San Francisco called Butterfly Restaurant.\n",
      "Embarcadero \t They do not have live music. They are at 33 The Embarcadero.\n",
      "San Francisco \t There is a nice restaurant in San Francisco called Chaat Corner.\n",
      "San Francisco \t Please confirm you want a reservation for 2 people at Chaat Corner in San Francisco today at 7:30 pm.\n",
      "San Francisco \t Please confirm you want the reservation in San Francisco on March 14th.\n",
      "Redwood City \t I'm looking for Mediterranean in Redwood City.\n",
      "Redwood City \t There is only 1 restaurant I could find in Redwood City, name Rocknwraps And Kabobs.\n",
      "Redwood City \t Ok, to confirm; you would like to eat at Rocknwraps And Kabobs in Redwood City at 5:30 pm today. I should make that reservation for 2 people?\n",
      "Fairfield \t I would love to have some kind of American food. Fairfield would be suitable to me.\n",
      "Fairfield \t Give me a moment ... There is a top-rated American restaurant in Fairfield and the restaurant name is Chuck E. Cheese's. Would you prefer this one?\n",
      "Fairfield \t Yup! Here, you go!!! Red Lobster, an American casual dining restaurant is nice one located in Fairfield. Do you like to have your food here?\n",
      "Texas \t Fortunately I can find one more for you. Texas Roadhouse, an American restaurant specializing in steaks is a very nice restaurant located in Fairfield.\n",
      "Roadhouse \t Fortunately I can find one more for you. Texas Roadhouse, an American restaurant specializing in steaks is a very nice restaurant located in Fairfield.\n",
      "Fairfield \t Fortunately I can find one more for you. Texas Roadhouse, an American restaurant specializing in steaks is a very nice restaurant located in Fairfield.\n",
      "Texas \t Please confirm me that you are booking a table at Texas Roadhouse located in Fairfield for 2 people today at 11:15 am. Is that right?\n",
      "Roadhouse \t Please confirm me that you are booking a table at Texas Roadhouse located in Fairfield for 2 people today at 11:15 am. Is that right?\n",
      "Fairfield \t Please confirm me that you are booking a table at Texas Roadhouse located in Fairfield for 2 people today at 11:15 am. Is that right?\n",
      "Oakland \t I want a economical place that serves alcohol in Oakland.\n",
      "Oakland \t I found 1 Umami Burger Oakland restaurant in Oakland.\n",
      "Oakland \t A table at Umami Burger Oakland, in Oakland, at 6 pm, on March 11th, for 2 people.\n",
      "Oakland \t I would like to find a restaurant in Oakland\n",
      "Oakland \t There is a restaurant called Alamar Kitchen & Bar in Oakland that meets your preferences\n",
      "Oakland \t So to confirm, you would like a 6:15 pm reservation for 2 today at Alamar Kitchen & Bar in Oakland.\n",
      "Oakland \t To confirm, you would like a reservation for 2 for 5:30 pm today at Alamar Kitchen & Bar in Oakland\n",
      "Santa Clara \t Look in Santa Clara. Any Southern restaurants?\n",
      "Santa Clara \t I found 4 matches. There's Bennigan's in Santa Clara.\n",
      "Santa Clara \t Please confirm: a table for 2 at Bennigan's in Santa Clara at 6:45 pm today\n",
      "Burlingame \t I want in Burlingame.\n",
      "Burlingame \t Sure. I found 4 restaurants. In Burlingame 37 North is a nice.\n",
      "Burlingame \t Please confirm. Booking a table in Burlingame 37 North restaurant today at 1:15 pm for 2 people.\n",
      "Burlingame \t Sorry, I was unable to make reservation. The phone number is 650-373-2237 and they don't have live music. Shall i book a table in Burlingame 37 North restaurant today at 1:30 pm for 2 people?\n",
      "Pleasanton \t Italian food in Pleasanton.\n",
      "Pleasanton \t I found De La Torre's Trattoria in Pleasanton.\n",
      "Pleasanton \t Forno Vecchio in Pleasanton.\n",
      "Pleasanton \t Gay 90's Pizza Co. in Pleasanton.\n",
      "Pleasanton \t Please confirm table at Gay 90's Pizza Co. for 2 in Pleasanton at 1:15 pm on March 10th.\n",
      "San Francisco \t There is a nice restaurant named Alice's in San Francisco\n",
      "San Francisco \t I've found 9 restaurants in San Francisco, Bamboo restaurant is a nice one.\n",
      "San Francisco \t Bei Fang Style is another nice restaurant in San Francisco\n",
      "San Francisco \t Please confirm if the following is correct: reservation at Bei Fang Style in San Francisco for 3, today at 1:30 pm.\n",
      "Oakland \t I'd like to find a place to eat in Oakland.\n",
      "Oakland \t I have found 2 different options. Cafe Eritrea D'afrique is a great choice in Oakland.\n",
      "Oakland \t I have found 1 place called Miliki Restaurant in Oakland.\n",
      "Oakland \t Please confirm the following: You want a table at Miliki Restaurant in Oakland at 11:15 am for 2 people on March 12th.\n",
      "Oakland \t Please confirm the following: You want a table for 2 people at Miliki Restaurant in Oakland at 11:15 am on the day after tomorrow.\n",
      "Walnut Creek \t I'm in Walnut Creek at the moment, please search there.\n",
      "Walnut Creek \t Okay, I found 3 restaurants in Walnut Creek. How about Panera Bread?\n",
      "Walnut Creek \t Okay sure, please confirm that you want a table for 2 at Panera Bread in Walnut Creek from 6 pm today.\n",
      "San Mateo \t I need it in San Mateo\n",
      "San Mateo \t Central Park Bistro is a good place in San Mateo\n",
      "San Mateo \t Book a table at Central Park Bistro in San Mateo for 5:45 pm for 2 people today.\n",
      "Newark \t I want some Indian food, preferable curry. Is there any place in Newark.\n",
      "Newark \t There are 4 places available. Biryani Pot in Newark has an excellent cuisine.\n",
      "Newark \t I would like to confirm the following: you would like to reserve a table for 2 persons for today at 7:30 pm at the Biryani Pot in Newark.\n",
      "San Bruno \t Somewhere to eat in San Bruno. Some Latin American food is what I'm looking for.\n",
      "San Bruno \t Celia's Mexican Restaurant is a nice restaurant located in San Bruno.\n",
      "San Bruno \t I found 1 restaurants. Chili's Grill & Bar is a nice restaurant located in San Bruno.\n",
      "San Bruno \t Please confirm the following details: Booking a table at Chili's Grill & Bar. The city is San Bruno. The reservation is at 5:30 pm. The date is March 9th. The reservation is for 6 people.\n",
      "San Bruno \t Please confirm the following details: The city is San Bruno. The reservation is at 6:30 pm.\n",
      "San Bruno \t Please confirm the following details: Booking a table at Chili's Grill & Bar. The city is San Bruno. The reservation is at 6:30 pm. The date is March 9th. The reservation is for 2 people.\n",
      "Berkeley \t I want Oriental food in Berkeley.\n",
      "Berkeley \t Mount Everest Restaurant in Berkeley is the number 1 hit.\n",
      "Berkeley \t I have one reservation at Mount Everest Restaurant in Berkeley on March 13th at 12:30 pm for 1 person. Is that correct?\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SF \t I want to eat in SF.\n",
      "San Francisco \t Okay, I've found 10 restaurants that could be good. How does Alice's in San Francisco sound?\n",
      "San Francisco \t Sure thing, please confirm: reservation for 2 people at Alice's in San Francisco at 11:45 am today.\n",
      "Morgan Hill \t I want some ramen, I'm really craving it. Can you find me an afforadable place in Morgan Hill?\n",
      "Morgan Hill \t Hanami Sushi does ramen in Morgan Hill. Is that okay?\n",
      "Nara \t Sushi Nara is another place that does ramen in Morgan Hill.\n",
      "Morgan Hill \t Sushi Nara is another place that does ramen in Morgan Hill.\n",
      "Fremont \t The location is Fremont, please.\n",
      "Fremont \t I found 4 restaurants in Fremont. Aniki's Sushi is a nice one.\n",
      "Fremont \t Please confirm that you'd like a table for 1 today at 12:30 pm at Aniki's Sushi in Fremont.\n",
      "San Mateo \t Please search in San Mateo, price is not an issue.\n",
      "San Mateo \t I have found a restaurant in San Mateo called Izakaya Ginji.\n",
      "Izakaya Ginji \t I have found a restaurant in San Mateo called Izakaya Ginji.\n",
      "Izakaya Ginji \t Can you confirm the following details: Reservation at Izakaya Ginji in San Mateo, for 1:30 pm today for 1 person.\n",
      "San Mateo \t Can you confirm the following details: Reservation at Izakaya Ginji in San Mateo, for 1:30 pm today for 1 person.\n",
      "South San Francisco \t I want to find one in South San Francisco.\n",
      "South San Francisco \t There are 3 restaurants. Amami San is a restaurant in South San Francisco.\n",
      "San Mateo \t Can you find another restaurant? I am looking for some thing with afforadable prices in San Mateo.\n",
      "San Mateo \t I have found 7 restaurants. Daiki Sushi is a restaurant located in San Mateo.\n",
      "San Mateo \t Dash Japanese Tapas And Sushi is a restaurant located in San Mateo.\n",
      "San Mateo \t Please confirm: You would like to book a table for 2 at Dash Japanese Tapas And Sushi in San Mateo at 12:15 pm today.\n",
      "San Francisco \t There are 4 restaurants that serve Pakistani cuisine in San Francisco. Aslam's Rasoi is a nice restaurant.\n",
      "San Francisco \t Ok, I am booking a table at Aslam's Rasoi in San Francisco on March 14th at 5:30 pm for 2 people. Is that correct?\n",
      "San Francisco \t Ok, I changed your reservation at Aslam's Rasoi in San Francisco to 3 people. Is that correct?\n",
      "San Francisco \t The phone number for Aslam's Rasoi in San Francisco is 415-695-0599.\n",
      "San Jose \t I'd like it to be in San Jose\n",
      "San Jose \t I found 4 restaurants in the San Jose area. I would recommend 4th St Pizza Co\n",
      "San Jose \t I would recommend a nice restaurant in San Jose called Mimi's Cafe\n",
      "San Jose \t Is this correct? I'm booking a table for 4 people at Mimi's Cafe in San Jose at 6 pm for next Thursday.\n",
      "San Francisco \t In San Francisco.\n",
      "Amelie San Francisco \t I found 10 possible restaurants. One is Amelie San Francisco, in San Francisco.\n",
      "San Francisco \t I found 10 possible restaurants. One is Amelie San Francisco, in San Francisco.\n",
      "Amelie San Francisco \t Please confirm the following: You want a table for 2 at 11:15 am today at Amelie San Francisco in San Francisco.\n",
      "San Francisco \t Please confirm the following: You want a table for 2 at 11:15 am today at Amelie San Francisco in San Francisco.\n",
      "Amelie San Francisco \t Please confirm the following: You want a table for 2 at 6 pm today at Amelie San Francisco in San Francisco.\n",
      "San Francisco \t Please confirm the following: You want a table for 2 at 6 pm today at Amelie San Francisco in San Francisco.\n",
      "Pleasanton \t I'd like the restaurant to be in Pleasanton. I'm in the mood for some Taiwanese food.\n",
      "Pleasanton \t Great! Andy & Yu's is a great restaurant in Pleasanton. Does that interest you at all?\n",
      "Pleasanton \t No problem. China Lounge Restaurant & Bar in Pleasanton is a great place. Does that interest you at all?\n",
      "Pleasanton \t There is! I recommend Hunan Chef Chinese Restaurant in Pleasanton. Does that interest you at all?\n",
      "Pleasanton \t Perfect. Can you confirm the following details, please? You have a table for 2 people at Hunan Chef Chinese Restaurant in Pleasanton at 12 pm today.\n",
      "Izakaya \t I'd love some Izakaya type food.\n",
      "St. Helena \t I'd like something in a moderate price range. How about Gastrobar type food, in St. Helena.\n",
      "St. Helena \t Please confirm that you would like to book a table for 2 people at the Cook Tavern & Pizzeria, in St. Helena, at 12:30 pm today.\n",
      "San Francisco \t I'd like to find a Burger in San Francisco.\n",
      "San Francisco \t I found about 10 restaurants in San Francisco. What about 1760?\n",
      "San Francisco \t Please confirm: reservation for 2 people next Tuesday at 5 pm at 1760 restaurant in San Francisco.\n",
      "Dublin \t Can you find me somewhere to eat in Dublin?\n",
      "Dublin \t Here are 7 options, one of which is China Paradise Restaurant in Dublin.\n",
      "Dublin \t I am making a reservation for 4 at China Paradise Restaurant in Dublin next Thursday at 2 pm. Is that correct?\n",
      "Dublin \t I am making a reservation for 2 people at China Paradise Restaurant in Dublin next Thursday at 2 pm. Is that correct?\n",
      "Dublin \t I am making a reservation for 2 people at China Paradise Restaurant in Dublin on March 9th at 2 pm. Is that correct?\n",
      "Dublin \t I am making a reservation for 2 people at China Paradise Restaurant in Dublin on March 9th at 1:45 pm. Is that correct?\n",
      "Pleasant Hill \t Ya, I prefer Asian food which I am always dreaming to eat and please search for me in Pleasant Hill.\n",
      "Pleasant Hill \t China Chef Restaurant is best for Asian foods which is located in Pleasant Hill, can you confirm whether this is convenient to you?\n"
     ]
    }
   ],
   "source": [
    "import spacy\n",
    "\n",
    "nlp = spacy.load(\"en_core_web_md\")\n",
    "\n",
    "\n",
    "with open(\"/Users/tdubon/GitHub/Mastering-spaCy/Chapter10/data/Chapter10/data/utterances.txt\", \"r\") as utterances:\n",
    "    for utterance in utterances:\n",
    "        utterance = utterance.strip()\n",
    "        doc = nlp(utterance)\n",
    "        ents = doc.ents\n",
    "        if ents:\n",
    "            for ent in ents:\n",
    "                if ent.label_ == \"GPE\":\n",
    "                    print(ent.text, \"\\t\", utterance)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a27f46f6",
   "metadata": {},
   "source": [
    "### Extracting date entities"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "e562addb",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "11:30 am \t I will be eating there at 11:30 am so make it for then.\n",
      "11:30 am \t Can you please confirm that you want to book a table for 2 at 11:30 am at the Bird Dog restaurant in Palo Alto for today.\n",
      "today \t Can you please confirm that you want to book a table for 2 at 11:30 am at the Bird Dog restaurant in Palo Alto for today.\n",
      "afternoon \t at afternoon 1:30\n",
      "1:30 \t at afternoon 1:30\n",
      "March 11th \t Please confirm the following details: Booking a table at Olive Garden Italian Restaurant at Milpitas on March 11th reservation time is 1:30 pm, The reservation is for 2 people\n",
      "1:30 pm \t Please confirm the following details: Booking a table at Olive Garden Italian Restaurant at Milpitas on March 11th reservation time is 1:30 pm, The reservation is for 2 people\n",
      "twelve in the afternoon \t No, time change to twelve in the afternoon for one people\n",
      "12 pm \t Please confirm the following details: the reservation time is 12 pm, The reservation is for 1 people\n",
      "a great day \t You are welcome, have a great day\n",
      "the 2nd of this month \t Reserve the table for the 2nd of this month at 17:15.\n",
      "17:15 \t Reserve the table for the 2nd of this month at 17:15.\n",
      "tomorrow \t Please confirm that you want to reserve a table for 2 at the Addis Restaurant in Berkeley for tomorrow at 5:15 pm.\n",
      "5:15 pm \t Please confirm that you want to reserve a table for 2 at the Addis Restaurant in Berkeley for tomorrow at 5:15 pm.\n",
      "17:30 \t Can you check if a table is available at 17:30?\n",
      "tomorrow \t Okay.Please confirm that you want to reserve a table for 2 at the Addis Restaurant in Berkeley for tomorrow at 5:30 pm.\n",
      "5:30 pm \t Okay.Please confirm that you want to reserve a table for 2 at the Addis Restaurant in Berkeley for tomorrow at 5:30 pm.\n",
      "a great day \t Have a great day!\n",
      "6:30 pm \t Please confirm: Booking a table for 1 at P.f. Chang's at 6:30 pm today in Palo Alto.\n",
      "today \t Please confirm: Booking a table for 1 at P.f. Chang's at 6:30 pm today in Palo Alto.\n",
      "a good day \t Have a good day.\n",
      "evening 5:30 on \t That would work great. I would like to make a reservation there at evening 5:30 on March 14th.\n",
      "March 14th \t That would work great. I would like to make a reservation there at evening 5:30 on March 14th.\n",
      "March 14th \t Please confirm the following: You want to eat at Golden Wok in San Mateo on March 14th starting at 5:30 pm. The reservation would be for 2 people.\n",
      "5:30 pm \t Please confirm the following: You want to eat at Golden Wok in San Mateo on March 14th starting at 5:30 pm. The reservation would be for 2 people.\n",
      "today \t Is there anything else today?\n",
      "a great day \t Have a great day!\n",
      "the 12th \t On the 12th at six pm\n",
      "six pm \t On the 12th at six pm\n",
      "March 12th \t Please confirm: Booking a table for 2 at Isla Restaurant in San Bruno. It is on March 12th at 6 pm.\n",
      "6 pm \t Please confirm: Booking a table for 2 at Isla Restaurant in San Bruno. It is on March 12th at 6 pm.\n",
      "a great day \t Have a great day.\n",
      "March 10th \t Yes make it for March 10th.\n",
      "six in the evening \t Reserve for six in the evening.\n",
      "6 pm \t Please confirm Ariake in San Jose for 6 pm on March 10th for 2.\n",
      "March 10th \t Please confirm Ariake in San Jose for 6 pm on March 10th for 2.\n",
      "a good day \t Thanks and have a good day.\n",
      "12:30 \t That restaurant will work can you please make a reservation for 12:30.\n",
      "12:30 pm \t I will be booking a table at Antonella's Ristorante in San Jose. The reservation will be set at 12:30 pm for today. And it is for 2 people is this correct.\n",
      "today \t I will be booking a table at Antonella's Ristorante in San Jose. The reservation will be set at 12:30 pm for today. And it is for 2 people is this correct.\n",
      "1:30 pm \t I'll reach there at 1:30 pm.\n",
      "today \t Please confirm that you want to make a reservation at Gay 90's Pizza Co. which is located in Pleasanton for 2 people. The reservation is for today at 1:30 pm.\n",
      "1:30 pm \t Please confirm that you want to make a reservation at Gay 90's Pizza Co. which is located in Pleasanton for 2 people. The reservation is for today at 1:30 pm.\n",
      "7 pm \t No, change the time to 7 pm and for one person only.\n",
      "7 pm \t Alright, please confirm that the reservation is at 7 pm for 1 person.\n",
      "next friday \t No, change it on next friday.\n",
      "next Friday \t Sure. Please confirm that the date is now next Friday and for 1 person.\n",
      "a great day \t Have a great day.\n",
      "this evening \t Yes, please make reservation for one people this evening 4:45\n",
      "4:45 \t Yes, please make reservation for one people this evening 4:45\n",
      "today \t Please confirm your booking at mimi's cafe in Fairfield for 1 people today at 4:45 pm\n",
      "4:45 pm \t Please confirm your booking at mimi's cafe in Fairfield for 1 people today at 4:45 pm\n",
      "today \t Sorry, they are fully booked. Shall i try to make reservation for 1 people today at 5 pm? The price range is Moderate at mimi's cafe and their phone number is 707-421-0835\n",
      "5 pm \t Sorry, they are fully booked. Shall i try to make reservation for 1 people today at 5 pm? The price range is Moderate at mimi's cafe and their phone number is 707-421-0835\n",
      "a nice day \t have a nice day\n",
      "half past 5 \t Please make it for half past 5 in the evening.\n",
      "today \t You'd like a table for 2 today at Mimi's Cafe in Fairfield at 5:30 pm. Is this correct?\n",
      "5:30 pm \t You'd like a table for 2 today at Mimi's Cafe in Fairfield at 5:30 pm. Is this correct?\n",
      "March 9th \t Let's review those changes. You'd like the reservation on March 9th for 3. Have I got it right?\n",
      "Monday next week \t I'm sorry. I need to make it on Monday next week at half past 12 in the afternoon.\n",
      "3 next Monday \t You'd like your reservation for 3 next Monday at 12:30 pm. Is that correct?\n",
      "12:30 pm \t You'd like your reservation for 3 next Monday at 12:30 pm. Is that correct?\n",
      "a nice day \t You're welcome. Have a nice day!\n",
      "A quarter past 5 \t A quarter past 5 in the evening, please.\n",
      "5:15 pm \t So you'd like a reservation for 2 at 5:15 pm today, at Rasa Sayang in Cupertino?\n",
      "today \t So you'd like a reservation for 2 at 5:15 pm today, at Rasa Sayang in Cupertino?\n",
      "March 13th \t No, I'd like it on March 13th.\n",
      "5:15 pm \t So you'd like a reservation at 5:15 pm on March 13th?\n",
      "March 13th \t So you'd like a reservation at 5:15 pm on March 13th?\n",
      "12:30 pm \t So you'd like a reservation at 12:30 pm on March 13th?\n",
      "March 13th \t So you'd like a reservation at 12:30 pm on March 13th?\n",
      "a nice day \t Have a nice day.\n",
      "7:15 pm \t Please confirm your reservation for 2 people at Aux Delices in San Francisco at 7:15 pm today.\n",
      "today \t Please confirm your reservation for 2 people at Aux Delices in San Francisco at 7:15 pm today.\n",
      "6:30 pm \t Sure, please confirm your reservation at 6:30 pm for 3 people.\n",
      "today \t They are located at 2327 Polk Street. Unfortunately, I was unable to make the reservation for you. Would you prefer a reservation at Aux Delices today for 3 people at 7 pm instead?\n",
      "7 pm \t They are located at 2327 Polk Street. Unfortunately, I was unable to make the reservation for you. Would you prefer a reservation at Aux Delices today for 3 people at 7 pm instead?\n",
      "a great day \t No problem, have a great day ahead!\n",
      "1:30 pm \t Please confirm to book a table for 2 persons in Asian Pearl Seafood REstaurant in Fremont at 1:30 pm on today.\n",
      "today \t Please confirm to book a table for 2 persons in Asian Pearl Seafood REstaurant in Fremont at 1:30 pm on today.\n",
      "March 10th \t change it to March 10th.\n",
      "March 10th \t Please confirm to book for 2 persons on March 10th.\n",
      "a great day \t Have a great day.\n",
      "the 13th of March \t Yes please reserve me a table for the 13th of March.\n",
      "evening \t 5 o\\\"clock in the evening please.\n",
      "March 13th \t Making a dinner reservation for 2 people at Juanita & Maude on March 13th at 5 pm in Albany. Is that Correct?\n",
      "5 pm \t Making a dinner reservation for 2 people at Juanita & Maude on March 13th at 5 pm in Albany. Is that Correct?\n",
      "a great day \t Have a great day.\n",
      "5:30 pm \t Okay, got it. I will book a table at 8 Immortals Restaurant in San Francisco for 5:30 pm today, for 6 people.\n",
      "today \t Okay, got it. I will book a table at 8 Immortals Restaurant in San Francisco for 5:30 pm today, for 6 people.\n",
      "twelve pm \t I want a reservation at this restaurant. Can you make it for twelve pm? For four people.\n",
      "12 pm \t Certainly, please confirm: Reservation at 8 Immortals Restaurant in San Francisco is for 12 pm today for 4 people.\n",
      "today \t Certainly, please confirm: Reservation at 8 Immortals Restaurant in San Francisco is for 12 pm today for 4 people.\n",
      "a great day \t Have a great day\n",
      "six pm \t Please reserve it for six pm.\n",
      "6 pm \t Sure, please confirm that you'd like to book a table for 2 at Chop Bar in Oakland from 6 pm today.\n",
      "today \t Sure, please confirm that you'd like to book a table for 2 at Chop Bar in Oakland from 6 pm today.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "half past 5 \t I want it at half past 5 in the evening.\n",
      "today \t Just to confirm, you want a table today at A16 in Oakland for 2 people at 5:30 pm?\n",
      "5:30 pm \t Just to confirm, you want a table today at A16 in Oakland for 2 people at 5:30 pm?\n",
      "a good day \t Have a good day.\n",
      "evening \t 5 o\\\"clock in the evening.\n",
      "today \t Ozora Sushi in Concord today for 2 people at 5 pm.\n",
      "5 pm \t Ozora Sushi in Concord today for 2 people at 5 pm.\n",
      "a good day \t Have a good day.\n",
      "this morning \t Make a reservation this morning 11:15 for three people.\n",
      "11:15 am \t Just to confirm: I am booking a table for 3 today at 11:15 am at the Rohnert Park location of Kyoto.\n",
      "a great day \t Have a great day.\n",
      "the 13th of this month \t Please make the reservation for the 13th of this month at afternoon 12.\n",
      "afternoon 12 \t Please make the reservation for the 13th of this month at afternoon 12.\n",
      "March 13th \t Great, I have a dining reservation for 2 people, located at the Hunan Empire Restaurant in San Francisco, on March 13th at 12 pm. Does that sound correct?\n",
      "12 pm \t Great, I have a dining reservation for 2 people, located at the Hunan Empire Restaurant in San Francisco, on March 13th at 12 pm. Does that sound correct?\n",
      "2001 \t I've successfully made your reservation. The address for the restaurant is 2001 Union Street.\n",
      "a wonderful day \t Have a wonderful day.\n",
      "Early evening \t Early evening 5:45 pm.\n",
      "5:45 pm \t Early evening 5:45 pm.\n",
      "today \t Please confirm: You would like to book a table, today at 5:45 pm, for 4 at Estampas Peruanas restaurant in redwood City?\n",
      "5:45 pm \t Please confirm: You would like to book a table, today at 5:45 pm, for 4 at Estampas Peruanas restaurant in redwood City?\n",
      "a lovely day \t Have a lovely day.\n",
      "today \t Okay. I have an 11 AM reservation for 2 people today, at Ipoh Garden Malaysian Cuisine in Millbrae. Is that correct?\n",
      "a great day \t My pleasure. Have a great day.\n",
      "12:30 pm \t Okay, so it's a reservation at Lovina in Calistoga at 12:30 pm today for 2 people?\n",
      "today \t Okay, so it's a reservation at Lovina in Calistoga at 12:30 pm today for 2 people?\n",
      "12:30 pm \t Okay, so it's 1 person at 12:30 pm?\n",
      "12:45 pm \t Please confirm: Booking a table for 2 people at 12:45 pm for today at the Grand Harbor in Burlingame.\n",
      "today \t Please confirm: Booking a table for 2 people at 12:45 pm for today at the Grand Harbor in Burlingame.\n",
      "a great day \t Have a great day.\n",
      "tonight \t I am tired of eating my cooking. I want to eat out tonight.\n",
      "1914 \t It is located on 1914 contra costa boulevard\n",
      "8 in the night \t 8 in the night, for 1 person\n",
      "8 pm \t I would like to confirm the following information: reservations for 1 for 8 pm today at the Matsu Sushi Japanese Restaurant at Pleasant Hill.\n",
      "today \t I would like to confirm the following information: reservations for 1 for 8 pm today at the Matsu Sushi Japanese Restaurant at Pleasant Hill.\n",
      "today \t I would like to confirm the following information: a table for 2 today at 8 pm at the Matsu Sushi Japanese Restaurant in Pleasant Hill.\n",
      "8 pm \t I would like to confirm the following information: a table for 2 today at 8 pm at the Matsu Sushi Japanese Restaurant in Pleasant Hill.\n",
      "11:45 \t I would like a table for 11:45 in the morning.\n",
      "11:45 am \t I would like to confirm the following information: reservations for 2 at 11:45 am\n",
      "12 pm \t I apologies but I was not able to make the reservation. Should I try for 12 pm today for 2 persons at Matsu Sushi Japanese Restaurant?\n",
      "today \t I apologies but I was not able to make the reservation. Should I try for 12 pm today for 2 persons at Matsu Sushi Japanese Restaurant?\n",
      "7:30 pm \t Yes, please make the reservation for 7:30 pm.\n",
      "7:30 pm \t I have booked a table at Noodles And Company in San Ramon at 7:30 pm today for 2 people. Is that correct?\n",
      "today \t I have booked a table at Noodles And Company in San Ramon at 7:30 pm today for 2 people. Is that correct?\n",
      "a great day \t Have a great day!\n",
      "the 7th of this month \t Great, can you make a reservation on the 7th of this month?\n",
      "about 7:30 pm \t How about 7:30 pm.\n",
      "next Thursday \t Can you confirm the details for me? You're booking a table for 2 next Thursday at 7:30 pm at China Stix in Santa Clara.\n",
      "7:30 pm \t Can you confirm the details for me? You're booking a table for 2 next Thursday at 7:30 pm at China Stix in Santa Clara.\n",
      "a great day \t Have a great day.\n",
      "Today \t Hey! Today feeling so tired... Wish to have some food outside in a good restaurant... Can you find me one?\n",
      "quarter past 1 \t Probably at quarter past 1 in the afternoon.\n",
      "1:15 pm \t Okay! Confirm me the details that you wish to book a table at Andes Cafe San Mateo in San Mateo at 1:15 pm by today and the reservation would be for 2 persons, right?\n",
      "today \t Okay! Confirm me the details that you wish to book a table at Andes Cafe San Mateo in San Mateo at 1:15 pm by today and the reservation would be for 2 persons, right?\n",
      "a fantastic day \t Have a fantastic day.\n",
      "5 pm \t 5 pm please. There's 3 people in the group.\n",
      "today \t Okay. Please confirm that you'd like a booking for 3 at Hot Crab in Dublin today at 5 pm.\n",
      "5 pm \t Okay. Please confirm that you'd like a booking for 3 at Hot Crab in Dublin today at 5 pm.\n",
      "11:30 am \t Please fix it at 11:30 am\n",
      "11:30 am \t ok here are the details please check or review it . Table booked at Bistro Don Giovanni which is in city Napa. Reservation time is 11:30 am for 2 people and the date is today.\n",
      "today \t ok here are the details please check or review it . Table booked at Bistro Don Giovanni which is in city Napa. Reservation time is 11:30 am for 2 people and the date is today.\n",
      "11:45 \t sorry make the reservation time as Morning 11:45\n",
      "11:45 am \t ok, please once again check the details. Reservation time for the table is at 11:45 am and for 2 people.\n",
      "a great day \t Take care have a great day.\n",
      "tonight \t I feel like Cambodian tonight.\n",
      "6:30 pm \t 6:30 pm. Please make the reservation for four people.\n",
      "6:30 pm \t Sure. A reservation at Anh Hong in Milpitas for 4 at 6:30 pm today. Is that correct?\n",
      "today \t Sure. A reservation at Anh Hong in Milpitas for 4 at 6:30 pm today. Is that correct?\n",
      "a nice day \t Have a nice day.\n",
      "7 pm \t Make the reservation for 7 pm.\n",
      "today \t Alright, please confirm that you'd like me to book a table for 2 at Lemongrass Bistro in Martinez for today at 7 pm.\n",
      "7 pm \t Alright, please confirm that you'd like me to book a table for 2 at Lemongrass Bistro in Martinez for today at 7 pm.\n",
      "next Monday \t No, make it for 1 person on next Monday please.\n",
      "next Monday \t Okay, so you'd like a reservation for 1 person on next Monday, right?\n",
      "a nice day \t Have a nice day!\n",
      "3rd of March \t That is good a good choice. Please make a reservation for six people for 3rd of March.\n",
      "Twelve pm \t Reserve it for Twelve pm.\n",
      "12 pm \t So, a table reservation at Meiko Sushi in Pleasanton will be for 12 pm on the day after tomorrow for 6 people. Please confirm the aforementioned details.\n",
      "the day after \t So, a table reservation at Meiko Sushi in Pleasanton will be for 12 pm on the day after tomorrow for 6 people. Please confirm the aforementioned details.\n",
      "tomorrow \t So, a table reservation at Meiko Sushi in Pleasanton will be for 12 pm on the day after tomorrow for 6 people. Please confirm the aforementioned details.\n",
      "a good day \t Have a good day.\n",
      "6:45 in the evening \t I want to be at 6:45 in the evening\n",
      "6:45 pm \t Book a table at Andy & Yu's in Pleasanton at 6:45 pm today for 2 person is correct?\n",
      "today \t Book a table at Andy & Yu's in Pleasanton at 6:45 pm today for 2 person is correct?\n",
      "12:45 \t No make it please at 12:45 for 4 people\n",
      "12:45 pm \t Reservation at 12:45 pm for 4 people Is OK?\n",
      "today \t Sorry the reservation was unsuccessful.Price range is moderate and they dont serve alcohol.Andy & Yu's is a nice place shall I try to meke a reservation for 4 person today at 12:30 pm?\n",
      "12:30 pm \t Sorry the reservation was unsuccessful.Price range is moderate and they dont serve alcohol.Andy & Yu's is a nice place shall I try to meke a reservation for 4 person today at 12:30 pm?\n",
      "a great day \t Have a great day!\n",
      "7:15 pm \t Please can you book it for 7:15 pm.\n",
      "today \t OK so you would like to book a table for 2 today at Furusato in Campbell for 7:15 pm.\n",
      "7:15 pm \t OK so you would like to book a table for 2 today at Furusato in Campbell for 7:15 pm.\n",
      "March 14th \t No, I'd like to book for March 14th.\n",
      "March 14th \t So to confirm, you would like a table at Furusato on March 14th.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Quarter past 5 \t Quarter past 5 in the evening.\n",
      "5:15 pm \t Please confirm a table at Pampas in Palo Alto at 5:15 pm for 2 people today.\n",
      "today \t Please confirm a table at Pampas in Palo Alto at 5:15 pm for 2 people today.\n",
      "evening 6:15 \t No, change it to evening 6:15.\n",
      "6:15 pm \t Okay. A table at Pampas at 6:15 pm.\n",
      "a wonderful day \t Have a wonderful day.\n",
      "1875 \t 1875 South Bascom Avenue is their street address.\n",
      "5:30 pm \t 5:30 pm should be prime pasta time.\n",
      "Today \t Ok, Let's be clear: 2 people for Dicicco's Ristorante Italiano Autentico. Today at 5:30 pm. in Campbell.\n",
      "5:30 pm \t Ok, Let's be clear: 2 people for Dicicco's Ristorante Italiano Autentico. Today at 5:30 pm. in Campbell.\n",
      "afternoon \t Well, it works for me, can you make a reservation for 4 people at 1 o\\\"clock in the afternoon?\n",
      "1 pm \t So I'll book a table for 4 at Anjappar Chettinad Restaurant in Milpitas for 1 pm today. Is that okay?\n",
      "today \t So I'll book a table for 4 at Anjappar Chettinad Restaurant in Milpitas for 1 pm today. Is that okay?\n",
      "evening 7:45 \t No, actually I want it for 3 people, and at evening 7:45.\n",
      "7:45 pm \t I've updated the reservation for 3 people, at 7:45 pm. Is that okay?\n",
      "a good day \t Have a good day!\n",
      "11 am \t The reservation needs to be at 11 am.\n",
      "today \t The reservation is for today at Red Chillies The Malabar Cuisine at 11 am in Milpitas. The reservation is for 2 people. Is this correct?\n",
      "11 am \t The reservation is for today at Red Chillies The Malabar Cuisine at 11 am in Milpitas. The reservation is for 2 people. Is this correct?\n",
      "March 14th \t The reservation should be for March 14th.\n",
      "March 14th \t The reservation is in Milpitas on March 14th. Is this correct?\n",
      "18:00 \t Please make the reservation for 18:00.\n",
      "March 14th \t The reservation is for March 14th and 6 pm. Is this correct?\n",
      "6 pm \t The reservation is for March 14th and 6 pm. Is this correct?\n",
      "the 7th of March \t I want it on the 7th of March at 13:00.\n",
      "13:00 \t I want it on the 7th of March at 13:00.\n",
      "next Thursday \t Your reservation will be for 1 person next Thursday at 1 pm at China Paradise Restaurant in Dublin. Is that correct?\n",
      "1 pm \t Your reservation will be for 1 person next Thursday at 1 pm at China Paradise Restaurant in Dublin. Is that correct?\n",
      "the 8th of March \t Sorry, it's actually for the 8th of March.\n",
      "next Friday \t So China Paradise Restaurant next Friday, is that right?\n",
      "11 am \t Okay, making a reservation for 2 people Shree Indian Cuisine in Petaluma at 11 am today. Is that correct?\n",
      "today \t Okay, making a reservation for 2 people Shree Indian Cuisine in Petaluma at 11 am today. Is that correct?\n",
      "a great day \t Have a great day!\n",
      "quarter past 5 \t Please reserve at quarter past 5 in the evening.\n",
      "5:15 pm \t Let me clarify your plans: You want a table at 54 Mint Ristorante Italiano in San Francisco at 5:15 pm. today for 2?\n",
      "today \t Let me clarify your plans: You want a table at 54 Mint Ristorante Italiano in San Francisco at 5:15 pm. today for 2?\n",
      "5:30 pm \t I apologize, I was unable to process this. how about trying a reservation for 54 Mint Ristorante Italiano for 2 today at 5:30 pm instead?\n",
      "a good day \t Have a good day.\n",
      "this afternoon \t How about this afternoon 1:45.\n",
      "1:45 \t How about this afternoon 1:45.\n",
      "1:45 pm \t One moment. Please confirm a reservation for 2 people, at Tabla Indian Restaurant in Foster City. At 1:45 pm today. Does this sound right?\n",
      "today \t One moment. Please confirm a reservation for 2 people, at Tabla Indian Restaurant in Foster City. At 1:45 pm today. Does this sound right?\n",
      "5:45 pm \t No it doesn't. Change to 5:45 pm on the 4th of this month please.\n",
      "5:45 pm \t One moment. Just to confirm: Your reservation is for 5:45 pm for next Monday. Does this sound right?\n",
      "next Monday \t One moment. Just to confirm: Your reservation is for 5:45 pm for next Monday. Does this sound right?\n",
      "a great day \t Your welcome. Have a great day.\n",
      "6 in the evening \t 6 in the evening.\n",
      "today \t Ok, I will reserve a table for 4 at Hunan Chili in Mountain View. The reservation is today at 6 pm.\n",
      "6 pm \t Ok, I will reserve a table for 4 at Hunan Chili in Mountain View. The reservation is today at 6 pm.\n",
      "a good day \t You're welcome. Have a good day.\n",
      "today \t Please confirm you are booking a table for 2 at Bamboo Restaurant, located in San Francisco today at 12 pm.\n",
      "12 pm \t Please confirm you are booking a table for 2 at Bamboo Restaurant, located in San Francisco today at 12 pm.\n",
      "a nice day \t Have a nice day.\n",
      "half past 7 \t The timing will be half past 7 in the evening.\n",
      "today \t Please confirm the details: You want me to make a reservation at Aq in San Francisco for today at 7:30 pm for 2 people.\n",
      "7:30 pm \t Please confirm the details: You want me to make a reservation at Aq in San Francisco for today at 7:30 pm for 2 people.\n",
      "a nice day \t I am glad I could help. Have a nice day.\n",
      "March 4th \t Ok. that is perfect for me and want to reserve that restaurant on March 4th.\n",
      "evening \t Do the reservation at evening 6:30\n",
      "6:30 \t Do the reservation at evening 6:30\n",
      "6:30 pm \t please give the confirmation details:Booking the table at Tokyo Village Grill & Sushi at Foster City and the reservation is for 2 people at 6:30 pm on next monday\n",
      "next monday \t please give the confirmation details:Booking the table at Tokyo Village Grill & Sushi at Foster City and the reservation is for 2 people at 6:30 pm on next monday\n",
      "a great Day \t have a great Day.\n",
      "6 in the evening \t Yes i wank to make a reservation at 6 in the evening.\n",
      "6 pm \t Ok Please check and verify the following details so that i will confirm. Table has been booked at 54 Mint Ristorante Italiano in the city San Francisco and the reservation time is at 6 pm for 2 people and the date is today.\n",
      "today \t Ok Please check and verify the following details so that i will confirm. Table has been booked at 54 Mint Ristorante Italiano in the city San Francisco and the reservation time is at 6 pm for 2 people and the date is today.\n",
      "a great day \t Take care have a great day\n",
      "12:15 \t Please make if at 12:15.\n",
      "today \t Please confirm the following details: Booking a table for 2 at Aappakadai in Pleasanton for today at 12:15 pm.\n",
      "12:15 pm \t Please confirm the following details: Booking a table for 2 at Aappakadai in Pleasanton for today at 12:15 pm.\n",
      "12 pm \t Sorry, I was unable to make the reservation. The phone number is 925-750-7709 and the address is 4555 Hopyard Road. Aappakadai is a nice restaurant. Shall I try for a reservation for 2 at 12 pm today?\n",
      "today \t Sorry, I was unable to make the reservation. The phone number is 925-750-7709 and the address is 4555 Hopyard Road. Aappakadai is a nice restaurant. Shall I try for a reservation for 2 at 12 pm today?\n",
      "a good day \t Have a good day.\n",
      "the evening \t Make it in the evening 7:15.\n",
      "today \t Just to confirm, you want a table at Gumba's in Sunnyvale for 2 people today at 7:15 pm?\n",
      "7:15 pm \t Just to confirm, you want a table at Gumba's in Sunnyvale for 2 people today at 7:15 pm?\n",
      "the afternoon \t No, make it in the afternoon 12:30.\n",
      "12:30 \t No, make it in the afternoon 12:30.\n",
      "12:30 pm \t Okay, so a reservation for 2 at 12:30 pm?\n",
      "next tuesday \t No, I want it to be for next tuesday, and only for one person.\n",
      "next tuesday \t Okay, a table for 1 next tuesday?\n",
      "a good day \t Have a good day.\n",
      "the evening \t I want it for the evening 4:45.\n",
      "4:45 \t I want it for the evening 4:45.\n",
      "today \t Okay, I'll you want a reservation at 1760 in San Francisco for 2 people, today at 4:45 pm, is that correct?\n",
      "4:45 pm \t Okay, I'll you want a reservation at 1760 in San Francisco for 2 people, today at 4:45 pm, is that correct?\n",
      "6 pm \t No, it's just for one person. And on second though, can you make it at 6 pm.\n",
      "6 pm \t So the reservation is for 1 person at 6 pm, correct?\n",
      "a quarter to 12 \t Yes. Make it for a quarter to 12 in the morning.\n",
      "today \t To confirm: You'd like me to make a reservation for today a 11:45 am for 2 guests at Rosie Mccann's Irish Pub & Restaurant in San Jose, correct?\n",
      "11:45 am \t To confirm: You'd like me to make a reservation for today a 11:45 am for 2 guests at Rosie Mccann's Irish Pub & Restaurant in San Jose, correct?\n",
      "11:30 am \t Please say yes to confirm the following. Booking a table at Contrasto in Oakland with a reservation at 11:30 am today. The reservation is a party of 2.\n",
      "today \t Please say yes to confirm the following. Booking a table at Contrasto in Oakland with a reservation at 11:30 am today. The reservation is a party of 2.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "a pleasant day \t Have a pleasant day.\n",
      "1:15 pm \t Confirm Bistro Aix and the city is San Francisco. Reservation os 1:15 pm and for 2 people today\n",
      "today \t Confirm Bistro Aix and the city is San Francisco. Reservation os 1:15 pm and for 2 people today\n",
      "today \t No live music at Bistro Aix but I shall try a reservation for 2 people today at 1:30 pm.\n",
      "1:30 pm \t No live music at Bistro Aix but I shall try a reservation for 2 people today at 1:30 pm.\n",
      "this evening \t Alright, please make a reservation for me at this evening 7:15.\n",
      "7:15 pm \t Please do confirm the following details: book seats at McDonald's in Dublin at 7:15 pm for 2 people today.\n",
      "today \t Please do confirm the following details: book seats at McDonald's in Dublin at 7:15 pm for 2 people today.\n",
      "7 pm \t Please make it at 7 pm.\n",
      "7 pm \t Please do confirm the following details: book seats at McDonald's in Dublin at 7 pm.\n",
      "14th of March \t Actually its for 3 people on 14th of March.\n",
      "March 14th \t Please do confirm the following details: book seats at McDonald's in Dublin for 3 people on March 14th.\n",
      "afternoon \t How about 12:30 in the afternoon?\n",
      "12:30 pm \t Please confirm your reservation. I have booked a table for 2 at La Paloma in Santa Clara at 12:30 pm today. Is that correct?\n",
      "today \t Please confirm your reservation. I have booked a table for 2 at La Paloma in Santa Clara at 12:30 pm today. Is that correct?\n",
      "this evening \t Great, I have booked your reservation. They will not have live music this evening, and the address is 2280 El Camino Real.\n",
      "12:45 pm \t Okay, that works. Please make me a reservation for 12:45 pm.\n",
      "12:45 pm \t So you'd like a 12:45 pm reservation today at Academy Bar and Kitchen in San Francisco for 2 people, is that right?\n",
      "today \t So you'd like a 12:45 pm reservation today at Academy Bar and Kitchen in San Francisco for 2 people, is that right?\n",
      "10:30 in the morning \t For two people, 10:30 in the morning.\n",
      "10:30 am \t Please confirm: Table for 2 at Lake Chalet, 10:30 am today in Oakland.\n",
      "today \t Please confirm: Table for 2 at Lake Chalet, 10:30 am today in Oakland.\n",
      "12:30 \t Try booking again, for 1 person at 12:30.\n",
      "12:30 pm \t Please confirm: Table for 1 at Lake Chalet, 12:30 pm today in Oakland.\n",
      "today \t Please confirm: Table for 1 at Lake Chalet, 12:30 pm today in Oakland.\n",
      "a nice day \t Have a nice day.\n",
      "afternoon \t I would like it at afternoon 1:15.\n",
      "1:15 pm \t Please confirm the following: You want a table at Angel Fish in Alameda at 1:15 pm today for 2 people.\n",
      "today \t Please confirm the following: You want a table at Angel Fish in Alameda at 1:15 pm today for 2 people.\n",
      "today \t Thanks so much. That's all for today.\n",
      "a wonderful day \t Have a wonderful day.\n",
      "tonight \t Hi I'm hungry and also need to let loose tonight. Can you find me a place to eat that serves alcohol?\n",
      "tonight \t I'm in the mood for some Chinese food tonight, preferably in Millbrae.\n",
      "evening 7:30 \t I'd like to be there at evening 7:30.\n",
      "today \t So you'd like a table for today at Tai Wu Restaurant for 2 people at 7:30 pm in Millbrae. Is that correct?\n",
      "7:30 pm \t So you'd like a table for today at Tai Wu Restaurant for 2 people at 7:30 pm in Millbrae. Is that correct?\n",
      "day \t Enjoy your day!\n",
      "the 4th of this month \t Make it on the 4th of this month please in the evening 6:30.\n",
      "the evening \t Make it on the 4th of this month please in the evening 6:30.\n",
      "6:30 \t Make it on the 4th of this month please in the evening 6:30.\n",
      "6:30 pm \t Please confirm your request: You'd like to book a table for 2 at Effie's Restaurant & Bar in Campbell at 6:30 pm for next Monday.\n",
      "next Monday \t Please confirm your request: You'd like to book a table for 2 at Effie's Restaurant & Bar in Campbell at 6:30 pm for next Monday.\n",
      "the 1st \t No I'd like a reservation on the 1st.\n",
      "today \t Please confirm the city you're booking in is Campbell and the date is for today.\n",
      "6:30 pm \t Please confirm: the reservation you want is for 6:30 pm for 3 people.\n",
      "a great day \t Have a great day.\n",
      "11:30 in the morning \t 11:30 in the morning seems good.\n",
      "11:30 AM \t Okay, please confirm the details one more time: Booking a table for 2 people at Crave Asia in Berkeley, the reservation should be for 11:30 AM and the date is today, correct?\n",
      "today \t Okay, please confirm the details one more time: Booking a table for 2 people at Crave Asia in Berkeley, the reservation should be for 11:30 AM and the date is today, correct?\n",
      "a great day \t Have a great day.\n",
      "half past 1 \t Yes. Reserve it for half past 1 in the afternoon.\n",
      "1:30 pm \t Please confirm that you are booking a table at risibisi in Petaluma. Your reservation will be at 1:30 pm. Set for today for 2 people.\n",
      "today \t Please confirm that you are booking a table at risibisi in Petaluma. Your reservation will be at 1:30 pm. Set for today for 2 people.\n",
      "a great day \t Have a great day.\n",
      "1300 \t There is a nice restaurant named 1300 On Fillmore located in San Francisco.\n",
      "11:00 \t I need it for 11:00.\n",
      "11 am \t Please confirm you want me to reserve a table for 2 people at 11 am at Aperto in San Francisco today.\n",
      "today \t Please confirm you want me to reserve a table for 2 people at 11 am at Aperto in San Francisco today.\n",
      "7:45 in the evening \t 7:45 in the evening please.\n",
      "7:45 pm \t Okay, please confirm that you'd like to book a table for 2 at P.F. Chang's in Pleasanton for 7:45 pm today.\n",
      "today \t Okay, please confirm that you'd like to book a table for 2 at P.F. Chang's in Pleasanton for 7:45 pm today.\n",
      "today \t Sorry, I couldn't book the table. How about booking the table at P.F. Chang's for 2, today at 7:30 pm instead?\n",
      "7:30 pm \t Sorry, I couldn't book the table. How about booking the table at P.F. Chang's for 2, today at 7:30 pm instead?\n",
      "today \t Okay, thanks. That's all for today.\n",
      "5 pm \t At 5 pm.\n",
      "5 pm \t You want to make a reservation at 5 pm today for 6 people at Brixx Pizzeria in Petaluma?\n",
      "today \t You want to make a reservation at 5 pm today for 6 people at Brixx Pizzeria in Petaluma?\n",
      "12:30 pm \t Sure! Please reserve a table for me at 12:30 pm.\n",
      "12:30 pm \t Please confirm your reservation for 2 at Meiko Sushi in Pleasanton at 12:30 pm today.\n",
      "today \t Please confirm your reservation for 2 at Meiko Sushi in Pleasanton at 12:30 pm today.\n",
      "7:15 pm \t Sure! Please confirm your reservation for 2 at Meiko Sushi in Pleasanton at 7:15 pm today.\n",
      "today \t Sure! Please confirm your reservation for 2 at Meiko Sushi in Pleasanton at 7:15 pm today.\n",
      "7:30 pm \t My apologies, I was not able to make the reservation. Alternatively, would you like a for 2 at Meiko Sushi in Pleasanton at 7:30 pm today instead? They are located at 4301 Valley Avenue B.\n",
      "today \t My apologies, I was not able to make the reservation. Alternatively, would you like a for 2 at Meiko Sushi in Pleasanton at 7:30 pm today instead? They are located at 4301 Valley Avenue B.\n",
      "a great day \t No problem! Have a great day ahead!\n",
      "the morning \t I'd like to reserve the table for lunch, so 11 o\\\"clock in the morning, for 2 people.\n",
      "today \t To review: You'd like to make an 11 am reservation at Amber restaurant in San Francisco for 2 people, today.\n",
      "today \t That's all I need today, thanks.\n",
      "a great day \t Have a great day.\n",
      "the 10th of this month \t Can you make it on the 10th of this month at a quarter to 7 in the evening?\n",
      "7 in the evening \t Can you make it on the 10th of this month at a quarter to 7 in the evening?\n",
      "March 10th \t Sure. Just to confirm, you'd like to make a reservation for 2 people at Ariake in San Jose on March 10th at 6:45 pm. Is that right?\n",
      "6:45 pm \t Sure. Just to confirm, you'd like to make a reservation for 2 people at Ariake in San Jose on March 10th at 6:45 pm. Is that right?\n",
      "afternoon \t Can you try making the reservation for 1 in the afternoon?\n",
      "March 10th \t Sure. Just to confirm, you'd like to make a reservation for 2 people at Ariake in San Jose on March 10th at 1 pm. Is that right?\n",
      "1 pm \t Sure. Just to confirm, you'd like to make a reservation for 2 people at Ariake in San Jose on March 10th at 1 pm. Is that right?\n",
      "Tuesday next week \t No, actually, change that to Tuesday next week.\n",
      "next Tuesday \t Ok, just to confirm, I'm making the reservation at Ariake for next Tuesday.\n",
      "a good day \t Have a good day.\n",
      "5:30 in \t 5:30 in the evening.\n",
      "Today \t Please confirm, Today at 5:30 pm in San Francisco, table 54 Mint Ristorante Italiano for 2 people.\n",
      "5:30 pm \t Please confirm, Today at 5:30 pm in San Francisco, table 54 Mint Ristorante Italiano for 2 people.\n",
      "March 3rd \t Please make it on March 3rd for 3 people.\n",
      "tomorrow \t It is ok, now verify for 3 people on day after tomorrow.\n",
      "11:30 am \t will you make it at 11:30 am on 7th of March?\n",
      "7th of March \t will you make it at 11:30 am on 7th of March?\n",
      "11:30 am \t The details of reservation are, time 11:30 am on next Thursday.\n",
      "next Thursday \t The details of reservation are, time 11:30 am on next Thursday.\n",
      "a wonderful day \t Have a wonderful day.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "March 7th \t Make it for one in the afternoon on March 7th.\n",
      "1 pm \t To confirm, the reservation is at Cala in San Francisco at 1 pm next Thursday for 2 people.\n",
      "next Thursday \t To confirm, the reservation is at Cala in San Francisco at 1 pm next Thursday for 2 people.\n",
      "a quarter past 12 \t Please change it to a quarter past 12 in the afternoon.\n",
      "12:15 pm \t To confirm, the reservation is at 12:15 pm for 2 people.\n",
      "next Thursday \t They do not have live music. I was unable to make the reservation at Cala, shall I try to reserve next Thursday at 12 pm for 2?\n",
      "12 pm \t They do not have live music. I was unable to make the reservation at Cala, shall I try to reserve next Thursday at 12 pm for 2?\n",
      "a good day \t Have a good day.\n",
      "1 pm \t At 1 pm, please.\n",
      "today \t Please confirm the following: You want a table for 3 at Olive Garden Italian Restaurant in Milpitas today at 1 pm?\n",
      "1 pm \t Please confirm the following: You want a table for 3 at Olive Garden Italian Restaurant in Milpitas today at 1 pm?\n",
      "today \t Please confirm the following details: You want a table for 4 at Olive Garden Italian Restaurant in Milpitas today at 1 pm?\n",
      "1 pm \t Please confirm the following details: You want a table for 4 at Olive Garden Italian Restaurant in Milpitas today at 1 pm?\n",
      "6:30 pm \t Alright, please confirm the following, you want a table at Olive Garden Italian Restaurant at 6:30 pm.\n",
      "evening \t Enjoy your evening!\n",
      "tomorrow \t That's okay, can you make me a reservation at this restaurant for day after tomorrow.\n",
      "11:30 in the morning \t 11:30 in the morning will be perfect.\n",
      "11:30 am on day \t Please confirm to book a table for 2 persons at Mcdonald's in San Jose at 11:30 am on day after tomorrow.\n",
      "tomorrow \t Please confirm to book a table for 2 persons at Mcdonald's in San Jose at 11:30 am on day after tomorrow.\n",
      "2nd of March \t Sorry, please change it to 3 persons and the date will be 2nd of March.\n",
      "tomorrow \t Please confirm to book the table for 3 persons on tomorrow.\n",
      "a great day \t Have a great day.\n",
      "20:00 \t Yes. I would like to make a reservation for 20:00.\n",
      "8 pm \t Great. Please verify that the following is correct: one table to Restaurant Chungdam in Santa Clara at 8 pm today for 2 people.\n",
      "today \t Great. Please verify that the following is correct: one table to Restaurant Chungdam in Santa Clara at 8 pm today for 2 people.\n",
      "7 in the evening \t Yes, I'd like to make the reservation at 7 in the evening.\n",
      "7 pm \t Great. Please verify that the following is correct: one table to Restaurant Chungdam in Santa Clara at 7 pm today for 2 people.\n",
      "today \t Great. Please verify that the following is correct: one table to Restaurant Chungdam in Santa Clara at 7 pm today for 2 people.\n",
      "day \t Enjoy the rest of your day!\n",
      "evening 5:45 \t Yes. Make it for evening 5:45.\n",
      "today \t Okay, please confirm: Table for 2 at Izakaya Sushi Ran in San Francisco today at 5:45 pm.\n",
      "5:45 pm \t Okay, please confirm: Table for 2 at Izakaya Sushi Ran in San Francisco today at 5:45 pm.\n",
      "the 8th \t No, it's for 1 person. Let's do it on the 8th.\n",
      "next Friday \t Okay, please confirm: reservation for 1 next Friday.\n",
      "11:15 am \t Please confirm: reservation for 3 people at 11:15 am.\n",
      "a great day \t Have a great day.\n",
      "today \t I have booked a table for 2 at B&b Kitchen & Wine Bar for today at 1:30 pm in Berkeley. Is this correct?\n",
      "1:30 pm \t I have booked a table for 2 at B&b Kitchen & Wine Bar for today at 1:30 pm in Berkeley. Is this correct?\n",
      "seven pm \t Make a seven pm reservation at a restaurant.\n",
      "today \t I have booked a table for 2 at B&b Kitchen & Wine Bar for today at 7 pm in Berkeley. Is this correct?\n",
      "7 pm \t I have booked a table for 2 at B&b Kitchen & Wine Bar for today at 7 pm in Berkeley. Is this correct?\n",
      "11:30 am \t 11:30 am is fine.\n",
      "today \t Please confirm: Making a reservation for 2 people at Celia's in San Mateo for today at 11:30 am.\n",
      "11:30 am \t Please confirm: Making a reservation for 2 people at Celia's in San Mateo for today at 11:30 am.\n",
      "evening 5:30 \t Yes. At evening 5:30, please.\n",
      "5:30 pm \t Please confirm Blue Mermaid in San Francisco at 5:30 pm for 2 people today.\n",
      "today \t Please confirm Blue Mermaid in San Francisco at 5:30 pm for 2 people today.\n",
      "a wonderful day \t Have a wonderful day.\n",
      "13:30 \t I'd like it to be at 13:30.\n",
      "today \t Okay, so you would like to reserve at table at Marhaba Indian & Pakistani Halal Cuisine in Oakland, for 2 people today at 1:30 pm. Is that correct?\n",
      "1:30 pm \t Okay, so you would like to reserve at table at Marhaba Indian & Pakistani Halal Cuisine in Oakland, for 2 people today at 1:30 pm. Is that correct?\n",
      "today \t Okay, so you'd like to make a reservation for 1 person today?\n",
      "a nice day \t Okay, have a nice day.\n",
      "1:30 in the afternoon \t 1:30 in the afternoon will be fine.\n",
      "1:30 pm \t Please confirm: Booking for a table for 4 persons at Soo Yuan Restaurant in Calistoga at 1:30 pm today.\n",
      "today \t Please confirm: Booking for a table for 4 persons at Soo Yuan Restaurant in Calistoga at 1:30 pm today.\n",
      "a great day \t Have a great day.\n",
      "12:15 in the afternoon \t It is for 1 person at 12:15 in the afternoon.\n",
      "12:15 pm \t You would like to book a table for 1 at 12:15 pm today at the Venus Restaurant in Berkeley. Is that correct?\n",
      "today \t You would like to book a table for 1 at 12:15 pm today at the Venus Restaurant in Berkeley. Is that correct?\n",
      "evening 6:45 \t Actually I changed my mind, make it at evening 6:45.\n",
      "6:45 pm \t You would like a reservation for 1 at 6:45 pm, is that correct?\n",
      "today \t I apologize but I was unable to confirm the reservation. You can reach them at 510-540-5950. Should I try a reservation for 1 person today at 7 pm at the Venus Restaurant?\n",
      "7 pm \t I apologize but I was unable to confirm the reservation. You can reach them at 510-540-5950. Should I try a reservation for 1 person today at 7 pm at the Venus Restaurant?\n",
      "today \t Is there anything else I can do for you today?\n",
      "a great day \t It was my pleasure. Have a great day.\n",
      "7 in the evening \t I would like to reserve for 7 in the evening for Monday next week.\n",
      "Monday next week \t I would like to reserve for 7 in the evening for Monday next week.\n",
      "7 pm \t Excellent. Can you make a reservation at Yet Wah Restaurant in San Rafael for 7 pm next Monday for 2 people?\n",
      "next Monday \t Excellent. Can you make a reservation at Yet Wah Restaurant in San Rafael for 7 pm next Monday for 2 people?\n",
      "tonight \t I would like to reserve a table for two people tonight.\n",
      "today \t Please verify the following details. You would like to reserve a table for 2 today at 6:30 pm at Mi Ranchito in Cotati.\n",
      "6:30 pm \t Please verify the following details. You would like to reserve a table for 2 today at 6:30 pm at Mi Ranchito in Cotati.\n",
      "Sunday this week \t Yes, I would like to make a reservation. Make a reservation for Sunday this week for 4 people.\n",
      "17:30 \t Mate it at 17:30.\n",
      "5:30 pm \t Please confirm the following details: You are booking a table at 54 Mint Ristorante Italiano, the city is San Francisco, the reservation is at 5:30 pm, the date is the day after tomorrow, and the reservation is for 4 people.\n",
      "the day after \t Please confirm the following details: You are booking a table at 54 Mint Ristorante Italiano, the city is San Francisco, the reservation is at 5:30 pm, the date is the day after tomorrow, and the reservation is for 4 people.\n",
      "tomorrow \t Please confirm the following details: You are booking a table at 54 Mint Ristorante Italiano, the city is San Francisco, the reservation is at 5:30 pm, the date is the day after tomorrow, and the reservation is for 4 people.\n",
      "Thursday next week \t No, I want it on Thursday next week.\n",
      "next Thursday \t Please confirm these updated details: The date is next Thursday, and the reservation is for 4 people.\n",
      "19:30 \t No make it at 19:30 and make the reservation for five poeple.\n",
      "7:30 pm \t Please confirm the updated details: the reservation is at 7:30 pm and it is for 5 people.\n",
      "a good day \t Have a good day.\n",
      "6 in the evening \t Let's do dinner at 6 in the evening\n",
      "6 pm \t Confirming your 6 pm booking today for 2 people at Baan Thai House & Wine Bar in San Francisco.\n",
      "today \t Confirming your 6 pm booking today for 2 people at Baan Thai House & Wine Bar in San Francisco.\n",
      "6 pm \t Confirming your 6 pm reservation for 1 then?\n",
      "today \t Please confirm: I am booking a table at Ming's, in San Bruno. The reservation is for 6 people, today at 12:30 pm.\n",
      "12:30 pm \t Please confirm: I am booking a table at Ming's, in San Bruno. The reservation is for 6 people, today at 12:30 pm.\n",
      "eleven am \t Make the reservation at eleven am.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "11 am \t Please confirm: I am booking a table at Ming's in San Bruno, the reservation is at 11 am today for 6 people.\n",
      "today \t Please confirm: I am booking a table at Ming's in San Bruno, the reservation is at 11 am today for 6 people.\n",
      "11 am \t Confirm: The reservation is at 11 am for 4 people.\n",
      "a good day \t Have a good day\n",
      "12:45 \t Please make if for 12:45.\n",
      "12:45 pm \t Please confirm the following: Booking a table for 2 today at Thai Time in the city of San Carlos for 12:45 pm.\n",
      "a good day \t Have a good day.\n",
      "5:15 in the evening \t Yes, please make a reservation for 5:15 in the evening, next Friday.\n",
      "next Friday \t Yes, please make a reservation for 5:15 in the evening, next Friday.\n",
      "next Friday \t Please confirm your reservation details: Reservation for 2 next Friday, at 5:15 PM, in Petaluma at the Flowering Tea House.\n",
      "5:15 PM \t Please confirm your reservation details: Reservation for 2 next Friday, at 5:15 PM, in Petaluma at the Flowering Tea House.\n",
      "the time quarter past 1 \t No, it's only for 1 person. Please make the time quarter past 1 in the afternoon.\n",
      "1:15 pm \t Please confirm your reservation details: Reservation for 1 person at 1:15 pm.\n",
      "the 13th of this month \t Actually, no. Make it the 13th of this month for 3 people.\n",
      "March 13th \t Please confirm your reservation details: March 13th for 3 people.\n",
      "March 13th \t Sorry, I was unable to process that request. Flowering Tea House does not serve alcohol. Shall I make a reservation for 3 people on March 13th at 1 pm?\n",
      "1 pm \t Sorry, I was unable to process that request. Flowering Tea House does not serve alcohol. Shall I make a reservation for 3 people on March 13th at 1 pm?\n",
      "a nice day \t Thank you, have a nice day.\n",
      "11:30 \t I want the reservation for 11:30.\n",
      "11:30 am \t Please confirm you want to make a reservation at 11:30 am today for 2 people at A Bellagio Italian Restaurant in Campbell.\n",
      "today \t Please confirm you want to make a reservation at 11:30 am today for 2 people at A Bellagio Italian Restaurant in Campbell.\n",
      "1:30 pm \t Can you confirm the reservation details please? I am booking a table at Bourbon Steak Restaurant in San Francisco. The reservation is for 1:30 pm for today and is for 2 people.\n",
      "today \t Can you confirm the reservation details please? I am booking a table at Bourbon Steak Restaurant in San Francisco. The reservation is for 1:30 pm for today and is for 2 people.\n",
      "a great day \t Be sure to have a great day!\n",
      "13:45 \t I want it to be at 13:45.\n",
      "1:45 pm \t Okay, I'll make you a reservation for 2 at Baci Bistro And Bar in Pleasanton at 1:45 pm today, is that correct?\n",
      "today \t Okay, I'll make you a reservation for 2 at Baci Bistro And Bar in Pleasanton at 1:45 pm today, is that correct?\n",
      "1:30 pm \t They're located at 500 Main Street, and they don't serve alcohol, sorry. I'm sorry but that time is all booked. Would you like a reservation for 2 at Baci Bistro And Bar for 1:30 pm today instead?\n",
      "today \t They're located at 500 Main Street, and they don't serve alcohol, sorry. I'm sorry but that time is all booked. Would you like a reservation for 2 at Baci Bistro And Bar for 1:30 pm today instead?\n",
      "afternoon 12 \t Make a reservation for one person at afternoon 12.\n",
      "today \t Please confirm: Making a reservation for 1 person today at 12 pm at 2g Japanese Brasserie in San Francisco.\n",
      "12 pm \t Please confirm: Making a reservation for 1 person today at 12 pm at 2g Japanese Brasserie in San Francisco.\n",
      "today \t Please confirm: Making a reservation for 3 people today at 12 pm at 2g Japanese Brasserie in San Francisco.\n",
      "12 pm \t Please confirm: Making a reservation for 3 people today at 12 pm at 2g Japanese Brasserie in San Francisco.\n",
      "12:30 in \t In second though I want it to be at 12:30 in the afternoon.\n",
      "today \t Making a reservation for 3 people today at 12:30 pm at 2g Japanese Brasserie in San Francisco.\n",
      "12:30 pm \t Making a reservation for 3 people today at 12:30 pm at 2g Japanese Brasserie in San Francisco.\n",
      "a good day \t Have a good day.\n",
      "4:30 pm \t I would love to make an appointment for one myself at 4:30 pm.\n",
      "today \t Just to be correct, you're booking a table at Okawa Japanese Restaurant today in Antioch and it will be for 1 your self at 4:30 pm.\n",
      "4:30 pm \t Just to be correct, you're booking a table at Okawa Japanese Restaurant today in Antioch and it will be for 1 your self at 4:30 pm.\n",
      "18:00 \t Instead of the previous time, can you make it at 18:00?\n",
      "6 pm \t Just to be correct, you're booking a table at Okawa Japanese Restaurant in Antioch at 6 pm. That will be today and that will be for 1.\n",
      "today \t Just to be correct, you're booking a table at Okawa Japanese Restaurant in Antioch at 6 pm. That will be today and that will be for 1.\n",
      "a great day \t Have a great day.\n",
      "1 pm \t Mcdonald's in Vallejo for 2 at 1 pm today. Correct?\n",
      "today \t Mcdonald's in Vallejo for 2 at 1 pm today. Correct?\n",
      "morning 10:30 \t Put it for morning 10:30.\n",
      "10:30 am \t Please verify the following information: Reserve a table for 2 people at Amelie San Francisco for 10:30 am today in San Francisco.\n",
      "today \t Please verify the following information: Reserve a table for 2 people at Amelie San Francisco for 10:30 am today in San Francisco.\n",
      "eleven in the morning \t I'd like to try making a reservation once more. Put it for eleven in the morning.\n",
      "11 am today \t Please verify the following information: reserving a table for 2 people at Amelie San Francisco at 11 am today in San Francisco.\n",
      "next Thursday \t Yes, please. I want to make a reservation for next Thursday.\n",
      "6:15 in the evening \t Please reserve me at 6:15 in the evening.\n",
      "next Thursday \t Please confirm these details: I'm reserving a table for 2 people at Red Tavern in San Francisco, for next Thursday at 6:15 pm.\n",
      "6:15 pm \t Please confirm these details: I'm reserving a table for 2 people at Red Tavern in San Francisco, for next Thursday at 6:15 pm.\n",
      "a good day \t Have a good day!\n",
      "4:45 \t evening 4:45 is perfect\n",
      "1760 \t booking table at 1760, city San francisco, time 4:45 pm, date today, reservation for 2 people. Confirm the details above\n",
      "4:45 pm \t booking table at 1760, city San francisco, time 4:45 pm, date today, reservation for 2 people. Confirm the details above\n",
      "today \t booking table at 1760, city San francisco, time 4:45 pm, date today, reservation for 2 people. Confirm the details above\n",
      "a great day \t Have a great day\n",
      "10:45 in the morning \t Please make it at 10:45 in the morning on the 4th of this month.\n",
      "10:45 am \t Please confirm the following: You want a table for 2 people at Osake in Santa Rosa at 10:45 am next Monday.\n",
      "next Monday \t Please confirm the following: You want a table for 2 people at Osake in Santa Rosa at 10:45 am next Monday.\n",
      "next Monday \t Unfortunately, the reservation failed. Their address is 2446 Patio Court and they don't have live music. Osake is a great choice. Shall I attempt to make a reservation for 2 next Monday at 11 am?\n",
      "11 am \t Unfortunately, the reservation failed. Their address is 2446 Patio Court and they don't have live music. Osake is a great choice. Shall I attempt to make a reservation for 2 next Monday at 11 am?\n",
      "a great day \t Have a great day!\n",
      "today \t Please confirm you want a reservation for 2 people at Chaat Corner in San Francisco today at 7:30 pm.\n",
      "7:30 pm \t Please confirm you want a reservation for 2 people at Chaat Corner in San Francisco today at 7:30 pm.\n",
      "March 14th \t No, I actually want to go on March 14th.\n",
      "March 14th \t Please confirm you want the reservation in San Francisco on March 14th.\n",
      "March 14th \t Please confirm you want a reservation for 3 people on March 14th.\n",
      "17:30 \t Yes, can you please make a reservation for 17:30?\n",
      "5:30 pm \t Ok, to confirm; you would like to eat at Rocknwraps And Kabobs in Redwood City at 5:30 pm today. I should make that reservation for 2 people?\n",
      "today \t Ok, to confirm; you would like to eat at Rocknwraps And Kabobs in Redwood City at 5:30 pm today. I should make that reservation for 2 people?\n",
      "the morning \t Please reserve in the morning at 11:15 am.\n",
      "11:15 am \t Please reserve in the morning at 11:15 am.\n",
      "today \t Please confirm me that you are booking a table at Texas Roadhouse located in Fairfield for 2 people today at 11:15 am. Is that right?\n",
      "11:15 am \t Please confirm me that you are booking a table at Texas Roadhouse located in Fairfield for 2 people today at 11:15 am. Is that right?\n",
      "13th of this month \t Nope, please book it on 13th of this month.\n",
      "March 13th \t Sure, I can do that. So, your booking would be at Texas Roadhouse on March 13th, right?\n",
      "a wonderful day \t Have a wonderful day.\n",
      "the 11th of March \t Sure! I want to make a reservation for the 11th of March.\n",
      "6 pm \t A table at Umami Burger Oakland, in Oakland, at 6 pm, on March 11th, for 2 people.\n",
      "March 11th \t A table at Umami Burger Oakland, in Oakland, at 6 pm, on March 11th, for 2 people.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "a good day \t Have a good day!\n",
      "6:15 pm \t I would like it for 6:15 pm\n",
      "6:15 pm \t So to confirm, you would like a 6:15 pm reservation for 2 today at Alamar Kitchen & Bar in Oakland.\n",
      "today \t So to confirm, you would like a 6:15 pm reservation for 2 today at Alamar Kitchen & Bar in Oakland.\n",
      "evening \t I would like to attempt a reservation for evening 5:30.\n",
      "5:30 \t I would like to attempt a reservation for evening 5:30.\n",
      "5:30 pm \t To confirm, you would like a reservation for 2 for 5:30 pm today at Alamar Kitchen & Bar in Oakland\n",
      "today \t To confirm, you would like a reservation for 2 for 5:30 pm today at Alamar Kitchen & Bar in Oakland\n",
      "a nice day \t Have a nice day.\n",
      "18:45 \t Do they have room for two at 18:45?\n",
      "6:45 pm \t Please confirm: a table for 2 at Bennigan's in Santa Clara at 6:45 pm today\n",
      "today \t Please confirm: a table for 2 at Bennigan's in Santa Clara at 6:45 pm today\n",
      "11:30 am \t No. Let's try 11:30 am on the 8th instead\n",
      "11:30 am \t Please confirm: 11:30 am next Friday?\n",
      "next Friday \t Please confirm: 11:30 am next Friday?\n",
      "1:15 pm \t Make it at 1:15 pm\n",
      "today \t Please confirm. Booking a table in Burlingame 37 North restaurant today at 1:15 pm for 2 people.\n",
      "1:15 pm \t Please confirm. Booking a table in Burlingame 37 North restaurant today at 1:15 pm for 2 people.\n",
      "today \t Sorry, I was unable to make reservation. The phone number is 650-373-2237 and they don't have live music. Shall i book a table in Burlingame 37 North restaurant today at 1:30 pm for 2 people?\n",
      "1:30 pm \t Sorry, I was unable to make reservation. The phone number is 650-373-2237 and they don't have live music. Shall i book a table in Burlingame 37 North restaurant today at 1:30 pm for 2 people?\n",
      "a great day \t Have a great day\n",
      "10th of this month \t Yes, please, on 10th of this month.\n",
      "1:15 pm \t Please confirm table at Gay 90's Pizza Co. for 2 in Pleasanton at 1:15 pm on March 10th.\n",
      "March 10th \t Please confirm table at Gay 90's Pizza Co. for 2 in Pleasanton at 1:15 pm on March 10th.\n",
      "a wonderful day \t Have a wonderful day.\n",
      "today \t Please confirm if the following is correct: reservation at Bei Fang Style in San Francisco for 3, today at 1:30 pm.\n",
      "1:30 pm \t Please confirm if the following is correct: reservation at Bei Fang Style in San Francisco for 3, today at 1:30 pm.\n",
      "a wonderful day \t Have a wonderful day.\n",
      "the 12th \t That works for me. I'd like to make a reservation on the 12th.\n",
      "11:15 in the morning \t Please schedule it at 11:15 in the morning.\n",
      "11:15 am \t Please confirm the following: You want a table at Miliki Restaurant in Oakland at 11:15 am for 2 people on March 12th.\n",
      "March 12th \t Please confirm the following: You want a table at Miliki Restaurant in Oakland at 11:15 am for 2 people on March 12th.\n",
      "the 3rd of this month \t I want to try booking again, but this time on the 3rd of this month.\n",
      "11:15 am \t Please confirm the following: You want a table for 2 people at Miliki Restaurant in Oakland at 11:15 am on the day after tomorrow.\n",
      "the day after \t Please confirm the following: You want a table for 2 people at Miliki Restaurant in Oakland at 11:15 am on the day after tomorrow.\n",
      "tomorrow \t Please confirm the following: You want a table for 2 people at Miliki Restaurant in Oakland at 11:15 am on the day after tomorrow.\n",
      "12:30 \t Never mind, make it for 12:30.\n",
      "12:30 pm \t Please confirm the following: You want a reservation at 12:30 pm for 2 people.\n",
      "today \t Thanks so much. That'll be all for today.\n",
      "a great day \t Have a great day!\n",
      "6 pm \t Okay sure, please confirm that you want a table for 2 at Panera Bread in Walnut Creek from 6 pm today.\n",
      "today \t Okay sure, please confirm that you want a table for 2 at Panera Bread in Walnut Creek from 6 pm today.\n",
      "5:45 in the evening \t I want it at 5:45 in the evening\n",
      "5:45 pm \t Book a table at Central Park Bistro in San Mateo for 5:45 pm for 2 people today.\n",
      "today \t Book a table at Central Park Bistro in San Mateo for 5:45 pm for 2 people today.\n",
      "today \t Central Park Bistro works and that will be for 2 people today at 6 pm\n",
      "6 pm \t Central Park Bistro works and that will be for 2 people today at 6 pm\n",
      "7:30 in the evening \t I prefer 7:30 in the evening.\n",
      "today \t I would like to confirm the following: you would like to reserve a table for 2 persons for today at 7:30 pm at the Biryani Pot in Newark.\n",
      "7:30 pm \t I would like to confirm the following: you would like to reserve a table for 2 persons for today at 7:30 pm at the Biryani Pot in Newark.\n",
      "Next Friday \t Yes I would. Next Friday for six people.\n",
      "5:30 in the evening \t I want it on the 9th at 5:30 in the evening.\n",
      "5:30 pm \t Please confirm the following details: Booking a table at Chili's Grill & Bar. The city is San Bruno. The reservation is at 5:30 pm. The date is March 9th. The reservation is for 6 people.\n",
      "March 9th \t Please confirm the following details: Booking a table at Chili's Grill & Bar. The city is San Bruno. The reservation is at 5:30 pm. The date is March 9th. The reservation is for 6 people.\n",
      "6:30 in the evening \t Change the time to 6:30 in the evening.\n",
      "6:30 pm \t Please confirm the following details: The city is San Bruno. The reservation is at 6:30 pm.\n",
      "6:30 pm \t Please confirm the following details: Booking a table at Chili's Grill & Bar. The city is San Bruno. The reservation is at 6:30 pm. The date is March 9th. The reservation is for 2 people.\n",
      "March 9th \t Please confirm the following details: Booking a table at Chili's Grill & Bar. The city is San Bruno. The reservation is at 6:30 pm. The date is March 9th. The reservation is for 2 people.\n",
      "a good day \t Have a good day.\n",
      "the 13th \t I want a reservation on the 13th at 12:30.\n",
      "12:30 \t I want a reservation on the 13th at 12:30.\n",
      "March 13th \t I have one reservation at Mount Everest Restaurant in Berkeley on March 13th at 12:30 pm for 1 person. Is that correct?\n",
      "12:30 pm \t I have one reservation at Mount Everest Restaurant in Berkeley on March 13th at 12:30 pm for 1 person. Is that correct?\n",
      "a nice day \t Have a nice day.\n",
      "the 1st of this month \t Yes, please. On the 1st of this month.\n",
      "around 11:45 \t For around 11:45.\n",
      "11:45 am \t Sure thing, please confirm: reservation for 2 people at Alice's in San Francisco at 11:45 am today.\n",
      "today \t Sure thing, please confirm: reservation for 2 people at Alice's in San Francisco at 11:45 am today.\n",
      "today \t I'm sorry, I was unable to make a reservation at that time. Shall I try for a reservation at Alice's for 2 people today at 12 pm? Their pricing is inexpensive and they do not have live music.\n",
      "12 pm \t I'm sorry, I was unable to make a reservation at that time. Shall I try for a reservation at Alice's for 2 people today at 12 pm? Their pricing is inexpensive and they do not have live music.\n",
      "a great day \t Have a great day.\n",
      "afternoon \t I'd like the table for afternoon 12:30 thanks.\n",
      "12:30 \t I'd like the table for afternoon 12:30 thanks.\n",
      "12:30 pm \t Please confirm that you'd like to book a table for 2 at Sushi Nara in Morgan Hill from 12:30 pm today.\n",
      "today \t Please confirm that you'd like to book a table for 2 at Sushi Nara in Morgan Hill from 12:30 pm today.\n",
      "the 12th \t No actually, I want it for three people on the 12th instead.\n",
      "3 on March 12th \t Okay sure, please confirm that you'd like a table for 3 on March 12th.\n",
      "next wednesday \t yes that works, place reservation next wednesday\n",
      "1 pm \t 1 pm\n",
      "1 pm \t Please confirm: A table at bajis cafe in mountain view at 1 pm on next wednesday for 2 people ?\n",
      "next wednesday \t Please confirm: A table at bajis cafe in mountain view at 1 pm on next wednesday for 2 people ?\n",
      "19:00 \t no make it 19:00 on march 9th\n",
      "7 pm march 9th \t Confirm: Reservation at 7 pm march 9th\n",
      "a great day \t no problem, have a great day.\n",
      "12:30 pm \t Please confirm that you'd like a table for 1 today at 12:30 pm at Aniki's Sushi in Fremont.\n",
      "the 14th of March \t No. Actually I need it for three people on the 14th of March.\n",
      "March 14th \t Please confirm that you need the reservation for 3 people on March 14th.\n",
      "1 pm \t No. Can you change it to 1 pm on the 9th?\n",
      "1 pm \t Please confirm that you want the reservation for 1 pm on March 9th.\n",
      "March 9th \t Please confirm that you want the reservation for 1 pm on March 9th.\n",
      "a good day \t Have a good day.\n",
      "1:30 pm \t I would like a reservation made for 1:30 pm. for 1 person please.\n",
      "1:30 pm \t Can you confirm the following details: Reservation at Izakaya Ginji in San Mateo, for 1:30 pm today for 1 person.\n",
      "today \t Can you confirm the following details: Reservation at Izakaya Ginji in San Mateo, for 1:30 pm today for 1 person.\n",
      "today \t Is their anything else I can help you with today?\n",
      "a nice day \t Have a nice day.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "12:15 \t Place a reservation for 12:15.\n",
      "12:15 pm \t Please confirm: You would like to book a table for 2 at Dash Japanese Tapas And Sushi in San Mateo at 12:15 pm today.\n",
      "today \t Please confirm: You would like to book a table for 2 at Dash Japanese Tapas And Sushi in San Mateo at 12:15 pm today.\n",
      "10:45 \t No it is for 1 person at 10:45.\n",
      "10:45 am \t Please confirm: The reservation is for 1 person at 10:45 am.\n",
      "today \t Sorry, I was unable to place a reservation at Dash Japanese Tapas And Sushi. Should I try one for 1 person today at 11 am?\n",
      "11 am \t Sorry, I was unable to place a reservation at Dash Japanese Tapas And Sushi. Should I try one for 1 person today at 11 am?\n",
      "a good day \t Have a good day.\n",
      "March 14th \t March 14th at 17:30, please.\n",
      "17:30 \t March 14th at 17:30, please.\n",
      "March 14th \t Ok, I am booking a table at Aslam's Rasoi in San Francisco on March 14th at 5:30 pm for 2 people. Is that correct?\n",
      "5:30 pm \t Ok, I am booking a table at Aslam's Rasoi in San Francisco on March 14th at 5:30 pm for 2 people. Is that correct?\n",
      "a good day \t You are welcome, have a good day.\n",
      "the 7th of March \t That's perfect! I need to book a reservation for 5 people on the 7th of March.\n",
      "6 pm \t Book it for four people at 6 pm\n",
      "6 pm \t Is this correct? I'm booking a table for 4 people at Mimi's Cafe in San Jose at 6 pm for next Thursday.\n",
      "next Thursday \t Is this correct? I'm booking a table for 4 people at Mimi's Cafe in San Jose at 6 pm for next Thursday.\n",
      "the 12th of this month \t No, I actually want it on the 12th of this month at twelve pm.\n",
      "twelve pm \t No, I actually want it on the 12th of this month at twelve pm.\n",
      "March 12th \t Is this correct? The reservation is on March 12th at 12 pm.\n",
      "12 pm \t Is this correct? The reservation is on March 12th at 12 pm.\n",
      "a great day \t Have a great day.\n",
      "about quarter past 11 \t How about quarter past 11 in the morning.\n",
      "11:15 am \t Please confirm the following: You want a table for 2 at 11:15 am today at Amelie San Francisco in San Francisco.\n",
      "today \t Please confirm the following: You want a table for 2 at 11:15 am today at Amelie San Francisco in San Francisco.\n",
      "6 pm \t Please confirm the following: You want a table for 2 at 6 pm today at Amelie San Francisco in San Francisco.\n",
      "today \t Please confirm the following: You want a table for 2 at 6 pm today at Amelie San Francisco in San Francisco.\n",
      "day \t Enjoy your day!\n",
      "12 in the afternoon \t 12 in the afternoon.\n",
      "12 pm \t Perfect. Can you confirm the following details, please? You have a table for 2 people at Hunan Chef Chinese Restaurant in Pleasanton at 12 pm today.\n",
      "today \t Perfect. Can you confirm the following details, please? You have a table for 2 people at Hunan Chef Chinese Restaurant in Pleasanton at 12 pm today.\n",
      "next Wednesday \t Okay, no problem. Can you confirm the following details, please? Your reservation is for 1 person next Wednesday.\n",
      "5:15 pm \t Oh, darn. My friends want dinner with me. Can we make it three people at 5:15 pm?\n",
      "5:15 pm \t Yes, that's no problem. Can you confirm the following details, please? Your reservation is for 3 people at 5:15 pm.\n",
      "a good day \t Have a good day!\n",
      "12:30 pm \t Please confirm that you would like to book a table for 2 people at the Cook Tavern & Pizzeria, in St. Helena, at 12:30 pm today.\n",
      "today \t Please confirm that you would like to book a table for 2 people at the Cook Tavern & Pizzeria, in St. Helena, at 12:30 pm today.\n",
      "17:00 \t No, make it for 4 people, at 17:00.\n",
      "5 pm \t Please confirm that the reservation is for 4 people at 5 pm.\n",
      "the 12th of March \t No. I want the reservation on the 12th of March, for 6 people.\n",
      "March 12th \t Please confirm the reservation is for 6 people, on March 12th.\n",
      "a nice day \t Have a nice day.\n",
      "Tuesday next week \t Yes, make it for Tuesday next week at evening 5.\n",
      "evening 5 \t Yes, make it for Tuesday next week at evening 5.\n",
      "next Tuesday \t Please confirm: reservation for 2 people next Tuesday at 5 pm at 1760 restaurant in San Francisco.\n",
      "5 pm \t Please confirm: reservation for 2 people next Tuesday at 5 pm at 1760 restaurant in San Francisco.\n",
      "a good day \t Have a good day.\n",
      "14:00 \t Please reserve the table for 14:00.\n",
      "next Thursday \t I am making a reservation for 4 at China Paradise Restaurant in Dublin next Thursday at 2 pm. Is that correct?\n",
      "2 pm \t I am making a reservation for 4 at China Paradise Restaurant in Dublin next Thursday at 2 pm. Is that correct?\n",
      "next Thursday \t I am making a reservation for 2 people at China Paradise Restaurant in Dublin next Thursday at 2 pm. Is that correct?\n",
      "2 pm \t I am making a reservation for 2 people at China Paradise Restaurant in Dublin next Thursday at 2 pm. Is that correct?\n",
      "March 9th \t Can you make a reservation at a restaurant for March 9th?\n",
      "March 9th \t I am making a reservation for 2 people at China Paradise Restaurant in Dublin on March 9th at 2 pm. Is that correct?\n",
      "2 pm \t I am making a reservation for 2 people at China Paradise Restaurant in Dublin on March 9th at 2 pm. Is that correct?\n",
      "1:45 \t Can you change it to afternoon 1:45?\n",
      "March 9th \t I am making a reservation for 2 people at China Paradise Restaurant in Dublin on March 9th at 1:45 pm. Is that correct?\n",
      "1:45 pm \t I am making a reservation for 2 people at China Paradise Restaurant in Dublin on March 9th at 1:45 pm. Is that correct?\n",
      "Tuesday next week \t Yes, that is fine, please make me a reservation on Tuesday next week.\n",
      "18:45 \t 18:45 will be fine.\n",
      "6:45 pm \t Please confirm: Booking a table for 2 persons at China Chef Restaurant located in Pleasant Hill at 6:45 pm on next Tuesday.\n",
      "next Tuesday \t Please confirm: Booking a table for 2 persons at China Chef Restaurant located in Pleasant Hill at 6:45 pm on next Tuesday.\n",
      "6:30 pm \t Sorry, I can't reserve at that time, can I book at 6:30 pm on next Tuesday for the same 2 people in China Chef Restaurant. The price is inexpensive, and they do not offer live music.\n",
      "next Tuesday \t Sorry, I can't reserve at that time, can I book at 6:30 pm on next Tuesday for the same 2 people in China Chef Restaurant. The price is inexpensive, and they do not offer live music.\n",
      "a great day \t Have a great day.\n"
     ]
    }
   ],
   "source": [
    "\n",
    "with open(\"/Users/tdubon/GitHub/Mastering-spaCy/Chapter10/data/Chapter10/data/utterances.txt\", \"r\") as utterances:\n",
    "    for utterance in utterances:\n",
    "        utterance = utterance.strip()\n",
    "        doc = nlp(utterance)\n",
    "        ents = doc.ents\n",
    "        if ents:\n",
    "            for ent in ents:\n",
    "                if ent.label_ == \"TIME\" or ent.label_ ==  \"DATE\":\n",
    "                    print(ent.text, \"\\t\", utterance)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7fd7505f",
   "metadata": {},
   "source": [
    "### Excluding results example"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "79cbc73d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "sentence = 'Have a nice day.'\n",
    "doc = nlp(sentence)\n",
    "wrong_matches = [\"a great day\", \"a nice day\"]\n",
    "date_ents = [ent for ent in doc.ents if ent.label_ == \"DATE\"] \n",
    "\n",
    "date_ents = list(filter(lambda e: e.text not in wrong_matches, date_ents))\n",
    "date_ents \n",
    "\n",
    " "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e027b79b",
   "metadata": {},
   "source": [
    "### using matcher to capture phone number entities"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "bf9bc40b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[The, number, is, 707, -, 766, -, 7600]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#investigate how the phone number is tokenized\n",
    "doc = nlp(\"The number is 707-766-7600\")\n",
    "[token for token in doc]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "37811e02",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Using pattern Matcher, we define how we want the number tokenized\n",
    "from spacy.matcher import Matcher\n",
    "matcher = Matcher(nlp.vocab)\n",
    "\n",
    "pattern = [{\"SHAPE\": \"ddd\"}, {\"TEXT\": \"-\"}, {\"SHAPE\": \"ddd\"}, {\"TEXT\": \"-\"}, {\"SHAPE\": \"dddd\"}] \n",
    "matcher.add(\"usPhoneNum\", [pattern])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "6c6647e1",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "707-766-7600\n"
     ]
    }
   ],
   "source": [
    "matches = matcher(doc)\n",
    "\n",
    "for mid, start, end in matches:\n",
    "    print(doc[start:end])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fa7eec4c",
   "metadata": {},
   "source": [
    "### Extracting cuisine types"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "e794f902",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "San Jose GPE\n",
      "Mexican NORP\n",
      "Italian NORP\n",
      "American NORP\n",
      "71 CARDINAL\n",
      "San Jose GPE\n",
      "71 CARDINAL\n",
      "North San Pedro Street LOC\n",
      "408 CARDINAL\n",
      "Bazille GPE\n",
      "San Jose GPE\n",
      "Palo Alto GPE\n",
      "7 CARDINAL\n",
      "Bird Dog ORG\n",
      "Palo Alto GPE\n",
      "11:30 am TIME\n",
      "2 CARDINAL\n",
      "11:30 am TIME\n",
      "Bird Dog ORG\n",
      "Palo Alto GPE\n",
      "today DATE\n",
      "420 Ramona Street FAC\n",
      "Punjabi NORP\n",
      "milpitas GPE\n",
      "Milpitas GPE\n",
      "Anjappar Chettinad Restaurant PERSON\n",
      "Milpitas Square FAC\n",
      "Aria Dining & Banquets Fine Indian Cuisine Restaurant ORG\n",
      "Milpitas GPE\n",
      "Restaurant ORG\n",
      "2 CARDINAL\n",
      "Italian Restaurant ORG\n",
      "Milpitas GPE\n",
      "11th ORDINAL\n",
      "afternoon 1:30 TIME\n",
      "Olive Garden Italian Restaurant FAC\n",
      "Milpitas GPE\n",
      "March 11th DATE\n",
      "1:30 pm TIME\n",
      "2 CARDINAL\n",
      "twelve in the afternoon TIME\n",
      "one CARDINAL\n",
      "12 pm TIME\n",
      "1 CARDINAL\n",
      "Restaurant ORG\n",
      "1350 CARDINAL\n",
      "Great Mall Drive FAC\n",
      "a great day DATE\n",
      "Ethiopian NORP\n",
      "Berkeley GPE\n",
      "4 CARDINAL\n",
      "Ethiopian NORP\n",
      "Berkeley GPE\n",
      "Addis Restaurant GPE\n",
      "4th ORDINAL\n",
      "the 2nd of this month DATE\n",
      "17:15 DATE\n",
      "2 CARDINAL\n",
      "the Addis Restaurant FAC\n",
      "Berkeley GPE\n",
      "tomorrow DATE\n",
      "5:15 pm TIME\n",
      "17:30 TIME\n",
      "2 CARDINAL\n",
      "the Addis Restaurant FAC\n",
      "Berkeley GPE\n",
      "tomorrow DATE\n",
      "5:30 pm TIME\n",
      "a great day DATE\n",
      "Palo Alto GPE\n",
      "4 CARDINAL\n",
      "Odori Japanese Cuisine ORG\n",
      "Palo Alto GPE\n",
      "P.f ORG\n",
      "Chang PERSON\n",
      "Palo Alto GPE\n",
      "evening TIME\n",
      "1 CARDINAL\n",
      "P.f ORG\n",
      "Chang PERSON\n",
      "6:30 pm TIME\n",
      "today DATE\n",
      "Palo Alto GPE\n",
      "650 CARDINAL\n",
      "a good day DATE\n",
      "Mexican NORP\n",
      "Italian NORP\n",
      "San Mateo GPE\n",
      "9 CARDINAL\n",
      "Andes Cafe San Mateo FAC\n",
      "San Mateo GPE\n",
      "650 CARDINAL\n",
      "Fuji Sukiyaki PERSON\n",
      "San Mateo GPE\n",
      "Golden Wok PERSON\n",
      "San Mateo GPE\n",
      "evening 5:30 on TIME\n",
      "March 14th DATE\n",
      "Golden Wok FAC\n",
      "San Mateo GPE\n",
      "March 14th DATE\n",
      "5:30 pm TIME\n",
      "2 CARDINAL\n",
      "today DATE\n",
      "a great day DATE\n",
      "Filipino NORP\n",
      "San Bruno GPE\n",
      "3 CARDINAL\n",
      "Cabalen PERSON\n",
      "San Bruno GPE\n",
      "Isla Restaurant ORG\n",
      "San Bruno GPE\n",
      "448 CARDINAL\n",
      "San Mateo Avenue GPE\n",
      "the 12th DATE\n",
      "six pm TIME\n",
      "2 CARDINAL\n",
      "Isla Restaurant FAC\n",
      "San Bruno GPE\n",
      "March 12th DATE\n",
      "6 pm TIME\n",
      "650 CARDINAL\n",
      "a great day DATE\n",
      "Indian NORP\n",
      "Chinese NORP\n",
      "San Jose GPE\n",
      "San Jose GPE\n",
      "Ariake GPE\n",
      "March 10th DATE\n",
      "six in the evening TIME\n",
      "Ariake PERSON\n",
      "San Jose GPE\n",
      "6 pm TIME\n",
      "March 10th DATE\n",
      "2 CARDINAL\n",
      "1008 CARDINAL\n",
      "a good day DATE\n",
      "Mexican NORP\n",
      "Chinese NORP\n",
      "San Jose GPE\n",
      "San Jose GPE\n",
      "Al Castello Ristorante FAC\n",
      "Angelo ORG\n",
      "Italian Restaurant ORG\n",
      "San Jose GPE\n",
      "408 CARDINAL\n",
      "436 Blossom Hill Road FAC\n",
      "Antonella PERSON\n",
      "San Jose GPE\n",
      "12:30 TIME\n",
      "Antonella's Ristorante ORG\n",
      "San Jose GPE\n",
      "12:30 pm TIME\n",
      "today DATE\n",
      "2 CARDINAL\n",
      "Pleasanton GPE\n",
      "Italian NORP\n",
      "Indian NORP\n",
      "Italian NORP\n",
      "5 CARDINAL\n",
      "De La Torre's ORG\n",
      "Pleasanton GPE\n",
      "6025 CARDINAL\n",
      "Forno Vecchio ORG\n",
      "n Pleasanton GPE\n",
      "Pleasanton GPE\n",
      "Gay 90's ORG\n",
      "Pizza Co. ORG\n",
      "1:30 pm TIME\n",
      "Gay 90's ORG\n",
      "Pizza Co. ORG\n",
      "Pleasanton GPE\n",
      "2 CARDINAL\n",
      "today DATE\n",
      "1:30 pm TIME\n",
      "7 pm TIME\n",
      "one CARDINAL\n",
      "7 pm TIME\n",
      "1 CARDINAL\n",
      "next friday DATE\n",
      "next Friday DATE\n",
      "1 CARDINAL\n",
      "288 Main Street FAC\n",
      "a great day DATE\n",
      "American NORP\n",
      "Italian NORP\n",
      "Fairfield GPE\n",
      "Mimi PERSON\n",
      "fairfield GPE\n",
      "one CARDINAL\n",
      "this evening TIME\n",
      "4:45 DATE\n",
      "Fairfield GPE\n",
      "1 CARDINAL\n",
      "today DATE\n",
      "4:45 pm TIME\n",
      "1 CARDINAL\n",
      "today DATE\n",
      "5 pm TIME\n",
      "707 CARDINAL\n",
      "a nice day DATE\n",
      "Breakfast ORG\n",
      "Fairfield GPE\n",
      "1 CARDINAL\n",
      "Mimi PERSON\n",
      "Fairfield GPE\n",
      "2 CARDINAL\n",
      "half past 5 DATE\n",
      "evening TIME\n",
      "2 CARDINAL\n",
      "today DATE\n",
      "Mimi's Cafe ORG\n",
      "Fairfield GPE\n",
      "5:30 pm TIME\n",
      "9th ORDINAL\n",
      "3 CARDINAL\n",
      "March 9th DATE\n",
      "3 CARDINAL\n",
      "Monday next week DATE\n",
      "afternoon TIME\n",
      "3 next Monday DATE\n",
      "12:30 pm TIME\n",
      "707 CARDINAL\n",
      "a nice day DATE\n",
      "Cupertino GPE\n",
      "American NORP\n",
      "Indian NORP\n",
      "Malaysian NORP\n",
      "2 CARDINAL\n",
      "Malaysian NORP\n",
      "Cupertino GPE\n",
      "Merlion Restaurant & Bar ORG\n",
      "Cupertino GPE\n",
      "A quarter past 5 DATE\n",
      "2 CARDINAL\n",
      "5:15 pm TIME\n",
      "today DATE\n",
      "Cupertino GPE\n",
      "March 13th DATE\n",
      "5:15 pm TIME\n",
      "March 13th DATE\n",
      "afternoon TIME\n",
      "12:30 pm TIME\n",
      "March 13th DATE\n",
      "a nice day DATE\n",
      "Italian NORP\n",
      "Indian NORP\n",
      "Cambodian NORP\n",
      "SF GPE\n",
      "Anh Hong GPE\n",
      "San Francisco GPE\n",
      "Aux Delices ORG\n",
      "San Francisco GPE\n",
      "2 CARDINAL\n",
      "Aux Delices FAC\n",
      "San Francisco GPE\n",
      "7:15 pm TIME\n",
      "today DATE\n",
      "3 CARDINAL\n",
      "6:30 pm TIME\n",
      "3 CARDINAL\n",
      "2327 Polk Street FAC\n",
      "Aux Delices FAC\n",
      "today DATE\n",
      "3 CARDINAL\n",
      "7 pm TIME\n",
      "a great day DATE\n",
      "Asian NORP\n",
      "Fremont City GPE\n",
      "3 CARDINAL\n",
      "Asian NORP\n",
      "Seafood Restaurant ORG\n",
      "Fremont city GPE\n",
      "2 CARDINAL\n",
      "Asian NORP\n",
      "Fremont GPE\n",
      "1:30 pm TIME\n",
      "today DATE\n",
      "March 10th DATE\n",
      "2 CARDINAL\n",
      "March 10th DATE\n",
      "a great day DATE\n",
      "Albany GPE\n",
      "Barbecue ORG\n",
      "2 CARDINAL\n",
      "Cafe Eugene PERSON\n",
      "Albany GPE\n",
      "Juanita & Maude ORG\n",
      "Albany GPE\n",
      "the 13th of March DATE\n",
      "5 CARDINAL\n",
      "evening TIME\n",
      "2 CARDINAL\n",
      "Juanita & Maude ORG\n",
      "March 13th DATE\n",
      "5 pm TIME\n",
      "Albany GPE\n",
      "a great day DATE\n",
      "Search San Francisco ORG\n",
      "Asian Fusion ORG\n",
      "8 CARDINAL\n",
      "San Francisco GPE\n",
      "415 CARDINAL\n",
      "six CARDINAL\n",
      "8 Immortals Restaurant ORG\n",
      "San Francisco GPE\n",
      "5:30 pm TIME\n",
      "today DATE\n",
      "6 CARDINAL\n",
      "twelve pm TIME\n",
      "four CARDINAL\n",
      "8 Immortals Restaurant FAC\n",
      "San Francisco GPE\n",
      "12 pm TIME\n",
      "today DATE\n",
      "4 CARDINAL\n",
      "a great day DATE\n",
      "Mexican NORP\n",
      "American NORP\n",
      "Sushi ORG\n",
      "Oakland GPE\n",
      "American NORP\n",
      "Chop Bar ORG\n",
      "Oakland GPE\n",
      "510 CARDINAL\n",
      "six pm TIME\n",
      "2 CARDINAL\n",
      "Chop Bar FAC\n",
      "Oakland GPE\n",
      "6 pm TIME\n",
      "today DATE\n",
      "247 4th Street FAC\n",
      "night TIME\n",
      "Oakland GPE\n",
      "Chinese NORP\n",
      "American NORP\n",
      "8 CARDINAL\n",
      "Oakland GPE\n",
      "two CARDINAL\n",
      "half past 5 DATE\n",
      "evening TIME\n",
      "today DATE\n",
      "A16 FAC\n",
      "Oakland GPE\n",
      "2 CARDINAL\n",
      "5:30 pm TIME\n",
      "a good day DATE\n",
      "Asian Fusion ORG\n",
      "Concord GPE\n",
      "Ozora Sushi PERSON\n",
      "Concord GPE\n",
      "5 CARDINAL\n",
      "evening TIME\n",
      "Ozora Sushi PERSON\n",
      "Concord GPE\n",
      "today DATE\n",
      "2 CARDINAL\n",
      "5 pm TIME\n",
      "785 CARDINAL\n",
      "a good day DATE\n",
      "Mexican NORP\n",
      "Rohnert Park GPE\n",
      "Chili's Grill & Bar ORG\n",
      "Rohnert Park GPE\n",
      "707 CARDINAL\n",
      "1 CARDINAL\n",
      "Kyoto GPE\n",
      "Rohnert Park GPE\n",
      "this morning TIME\n",
      "three CARDINAL\n",
      "3 CARDINAL\n",
      "11:15 am TIME\n",
      "Rohnert Park FAC\n",
      "Kyoto GPE\n",
      "5 Padre Parkway FAC\n",
      "a great day DATE\n",
      "San Fran GPE\n",
      "Golden Gate Indian Cuisine & Pizza ORG\n",
      "San Francisco GPE\n",
      "415 CARDINAL\n",
      "1 CARDINAL\n",
      "Hunan Empire Restaurant GPE\n",
      "San Francisco GPE\n",
      "two CARDINAL\n",
      "the 13th of this month DATE\n",
      "12 CARDINAL\n",
      "2 CARDINAL\n",
      "the Hunan Empire Restaurant FAC\n",
      "San Francisco GPE\n",
      "March 13th DATE\n",
      "12 pm TIME\n",
      "2001 DATE\n",
      "Union Street EVENT\n",
      "a wonderful day DATE\n",
      "Redwood City GPE\n",
      "Spanish NORP\n",
      "Redwood City GPE\n",
      "1 CARDINAL\n",
      "Spanish NORP\n",
      "Estampas Peruanas Restaurant ORG\n",
      "four CARDINAL\n",
      "Early evening TIME\n",
      "5:45 pm TIME\n",
      "today DATE\n",
      "5:45 pm TIME\n",
      "4 CARDINAL\n",
      "Estampas Peruanas ORG\n",
      "City GPE\n",
      "650 CARDINAL\n",
      "a lovely day DATE\n",
      "Singaporean NORP\n",
      "Millbrae GPE\n",
      "1 CARDINAL\n",
      "Ipoh Garden FAC\n",
      "Malaysian Cuisine ORG\n",
      "Millbrae GPE\n",
      "650 CARDINAL\n",
      "11 CARDINAL\n",
      "2 CARDINAL\n",
      "today DATE\n",
      "Ipoh Garden Malaysian Cuisine FAC\n",
      "Millbrae GPE\n",
      "a great day DATE\n",
      "Calistoga GPE\n",
      "Chinese NORP\n",
      "Mexican NORP\n",
      "Cambodian NORP\n",
      "Calistoga Thai Kitchen ORG\n",
      "Calistoga GPE\n",
      "Barbecue LOC\n",
      "1 CARDINAL\n",
      "Lovina GPE\n",
      "Calistoga GPE\n",
      "half past 12 CARDINAL\n",
      "two CARDINAL\n",
      "Lovina GPE\n",
      "Calistoga GPE\n",
      "12:30 pm TIME\n",
      "today DATE\n",
      "2 CARDINAL\n",
      "1 CARDINAL\n",
      "1 CARDINAL\n",
      "12:30 pm TIME\n",
      "Mexican NORP\n",
      "Chinese NORP\n",
      "Burlingame GPE\n",
      "2 CARDINAL\n",
      "One CARDINAL\n",
      "Grand Harbor LOC\n",
      "Burlingame GPE\n",
      "1 CARDINAL\n",
      "afternoon TIME\n",
      "2 CARDINAL\n",
      "12:45 pm TIME\n",
      "today DATE\n",
      "the Grand Harbor FAC\n",
      "Burlingame GPE\n",
      "650 CARDINAL\n",
      "a great day DATE\n",
      "tonight TIME\n",
      "Izakaya NORP\n",
      "Rohnert Park GPE\n",
      "Rohnert Park GPE\n",
      "Hana\n",
      " ORG\n",
      "101 CARDINAL\n",
      "Pleasant Hill FAC\n",
      "2 CARDINAL\n",
      "Pleasant Hill GPE\n",
      "Matsu Sushi Japanese Restaurant ORG\n",
      "1914 DATE\n",
      "costa boulevard\n",
      " ORG\n",
      "8 in the night TIME\n",
      "1 CARDINAL\n",
      "1 CARDINAL\n",
      "8 pm TIME\n",
      "today DATE\n",
      "the Matsu Sushi Japanese Restaurant FAC\n",
      "Pleasant Hill FAC\n",
      "2 CARDINAL\n",
      "2 CARDINAL\n",
      "today DATE\n",
      "8 pm TIME\n",
      "the Matsu Sushi Japanese Restaurant FAC\n",
      "Pleasant Hill GPE\n",
      "11:45 TIME\n",
      "morning TIME\n",
      "2 CARDINAL\n",
      "11:45 am TIME\n",
      "12 pm TIME\n",
      "today DATE\n",
      "2 CARDINAL\n",
      "Matsu Sushi Japanese Restaurant ORG\n",
      "day DATE\n",
      "Mexican NORP\n",
      "Italian NORP\n",
      "San Ramon GPE\n",
      "San Ramon GPE\n",
      "925 CARDINAL\n",
      "3141 Crow Canyon Place FAC\n",
      "7:30 pm TIME\n",
      "Noodles And Company ORG\n",
      "San Ramon GPE\n",
      "7:30 pm TIME\n",
      "today DATE\n",
      "2 CARDINAL\n",
      "a great day DATE\n",
      "Cupertino GPE\n",
      "Cupertino GPE\n",
      "A Plus Cafe WORK_OF_ART\n",
      "Santa Clara GPE\n",
      "Santa Clara GPE\n",
      "Chef Ming's ORG\n",
      "China Stix PERSON\n",
      "Santa Clara GPE\n",
      "the 7th of this month DATE\n",
      "about 7:30 pm TIME\n",
      "2 CARDINAL\n",
      "next Thursday DATE\n",
      "7:30 pm TIME\n",
      "China Stix FAC\n",
      "Santa Clara GPE\n",
      "a great day DATE\n",
      "Today DATE\n",
      "Wish PERSON\n",
      "Chinese NORP\n",
      "American NORP\n",
      "Italian NORP\n",
      "Seafood ORG\n",
      "San Mateo GPE\n",
      "9 CARDINAL\n",
      "Andes Cafe LOC\n",
      "San Mateo GPE\n",
      "San Mateo GPE\n",
      "2317 South El Camino Real FAC\n",
      "quarter past 1 DATE\n",
      "afternoon TIME\n",
      "Andes Cafe San Mateo FAC\n",
      "San Mateo GPE\n",
      "1:15 pm TIME\n",
      "today DATE\n",
      "2 CARDINAL\n",
      "a fantastic day DATE\n",
      "Dublin GPE\n",
      "2 CARDINAL\n",
      "One CARDINAL\n",
      "the Black Angus Steakhouse ORG\n",
      "Dublin GPE\n",
      "Hot Crab ORG\n",
      "Dublin GPE\n",
      "5 pm TIME\n",
      "3 CARDINAL\n",
      "3 CARDINAL\n",
      "Hot Crab ORG\n",
      "Dublin GPE\n",
      "today DATE\n",
      "5 pm TIME\n",
      "925-999-9580 CARDINAL\n",
      "4288 CARDINAL\n",
      "Indian NORP\n",
      "Chinese NORP\n",
      "Napa GPE\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1 CARDINAL\n",
      "Bistro Don Giovanni PERSON\n",
      "Napa GPE\n",
      "4110 CARDINAL\n",
      "Howard Lane PERSON\n",
      "2 CARDINAL\n",
      "11:30 am TIME\n",
      "Bistro Don Giovanni FAC\n",
      "Napa GPE\n",
      "11:30 am TIME\n",
      "2 CARDINAL\n",
      "today DATE\n",
      "Morning 11:45 TIME\n",
      "11:45 am TIME\n",
      "2 CARDINAL\n",
      "707 CARDINAL\n",
      "a great day DATE\n",
      "Milpitas GPE\n",
      "Mexican NORP\n",
      "American NORP\n",
      "Cambodian NORP\n",
      "tonight TIME\n",
      "Cambodian NORP\n",
      "Milpitas GPE\n",
      "Anh Hong GPE\n",
      "6:30 pm TIME\n",
      "four CARDINAL\n",
      "Anh Hong GPE\n",
      "Milpitas GPE\n",
      "4 CARDINAL\n",
      "6:30 pm TIME\n",
      "today DATE\n",
      "Anh Hong GPE\n",
      "a nice day DATE\n",
      "Martinez GPE\n",
      "Thai NORP\n",
      "1 CARDINAL\n",
      "Martinez GPE\n",
      "Lemongrass Bistro ORG\n",
      "925 CARDINAL\n",
      "7 pm TIME\n",
      "2 CARDINAL\n",
      "Lemongrass Bistro FAC\n",
      "Martinez GPE\n",
      "today DATE\n",
      "7 pm TIME\n",
      "1 CARDINAL\n",
      "next Monday DATE\n",
      "1 CARDINAL\n",
      "next Monday DATE\n",
      "501 CARDINAL\n",
      "a nice day DATE\n",
      "Italian NORP\n",
      "American NORP\n",
      "Pleasanton GPE\n",
      "Ramen GPE\n",
      "Meiko Sushi ORG\n",
      "Pleasanton GPE\n",
      "six CARDINAL\n",
      "3rd of March DATE\n",
      "Reservation ORG\n",
      "Twelve pm TIME\n",
      "Meiko Sushi ORG\n",
      "Pleasanton GPE\n",
      "12 pm TIME\n",
      "the day after DATE\n",
      "tomorrow DATE\n",
      "6 CARDINAL\n",
      "a good day DATE\n",
      "Pleasanton GPE\n",
      "Noodles PERSON\n",
      "Pleasanton GPE\n",
      "Andy & Yu's ORG\n",
      "925-750-8888 CARDINAL\n",
      "2 CARDINAL\n",
      "6:45 in the evening TIME\n",
      "Andy & Yu's ORG\n",
      "Pleasanton GPE\n",
      "6:45 pm TIME\n",
      "today DATE\n",
      "2 CARDINAL\n",
      "12:45 TIME\n",
      "4 CARDINAL\n",
      "12:45 pm TIME\n",
      "4 CARDINAL\n",
      "Liquor ORG\n",
      "Andy & Yu's PERSON\n",
      "4 CARDINAL\n",
      "today DATE\n",
      "12:30 pm TIME\n",
      "a great day DATE\n",
      "Barbecue LOC\n",
      "Campbell ORG\n",
      "3 CARDINAL\n",
      "Campbell ORG\n",
      "Effie ORG\n",
      "Restaurant & Bar ORG\n",
      "331 CARDINAL\n",
      "West Hacienda Avenue FAC\n",
      "2 CARDINAL\n",
      "Campbell ORG\n",
      "Campbell ORG\n",
      "Sushi ORG\n",
      "Furusato PERSON\n",
      "Campbell ORG\n",
      "7:15 pm TIME\n",
      "2 CARDINAL\n",
      "today DATE\n",
      "Furusato PERSON\n",
      "Campbell ORG\n",
      "7:15 pm TIME\n",
      "March 14th DATE\n",
      "March 14th DATE\n",
      "408 CARDINAL\n",
      "day DATE\n",
      "Palo Alto GPE\n",
      "Brazilian NORP\n",
      "Palo Alto GPE\n",
      "Quarter past 5 DATE\n",
      "evening TIME\n",
      "Palo Alto GPE\n",
      "5:15 pm TIME\n",
      "2 CARDINAL\n",
      "today DATE\n",
      "evening 6:15 TIME\n",
      "6:15 pm TIME\n",
      "650 CARDINAL\n",
      "a wonderful day DATE\n",
      "Chinese NORP\n",
      "Mexican NORP\n",
      "Campbell ORG\n",
      "5 CARDINAL\n",
      "Bellagio Italian Restaurant ORG\n",
      "Campbell ORG\n",
      "Buca Di Beppo PERSON\n",
      "Campbell ORG\n",
      "1875 DATE\n",
      "Bascom Avenue FAC\n",
      "Dicicco ORG\n",
      "Ristorante Italiano Autentico ORG\n",
      "Campbell ORG\n",
      "5:30 pm TIME\n",
      "2 CARDINAL\n",
      "Dicicco ORG\n",
      "Ristorante Italiano Autentico ORG\n",
      "Today DATE\n",
      "5:30 pm TIME\n",
      "Campbell ORG\n",
      "Milpitas GPE\n",
      "Indian NORP\n",
      "4 CARDINAL\n",
      "Anjappar Chettinad Restaurant PERSON\n",
      "Milpitas GPE\n",
      "408 CARDINAL\n",
      "4 CARDINAL\n",
      "1 CARDINAL\n",
      "afternoon TIME\n",
      "4 CARDINAL\n",
      "Anjappar Chettinad Restaurant FAC\n",
      "Milpitas GPE\n",
      "1 pm TIME\n",
      "today DATE\n",
      "3 CARDINAL\n",
      "evening 7:45 TIME\n",
      "3 CARDINAL\n",
      "7:45 pm TIME\n",
      "a good day DATE\n",
      "Milpitas GPE\n",
      "2 CARDINAL\n",
      "Milpitas GPE\n",
      "The Olive Garden Italian Restaurant ORG\n",
      "Milpitas GPE\n",
      "Red Chillies The Malabar Cuisine ORG\n",
      "167 CARDINAL\n",
      "South Main Street FAC\n",
      "Red Chillies ORG\n",
      "11 am TIME\n",
      "today DATE\n",
      "Red Chillies The Malabar Cuisine ORG\n",
      "11 am TIME\n",
      "Milpitas GPE\n",
      "2 CARDINAL\n",
      "March 14th DATE\n",
      "Milpitas GPE\n",
      "March 14th DATE\n",
      "18:00 DATE\n",
      "March 14th DATE\n",
      "6 pm TIME\n",
      "669 CARDINAL\n",
      "day DATE\n",
      "Italian NORP\n",
      "American NORP\n",
      "Noodles ORG\n",
      "Dublin GPE\n",
      "7 CARDINAL\n",
      "China Paradise Restaurant FAC\n",
      "Dublin GPE\n",
      "one CARDINAL\n",
      "the 7th of March DATE\n",
      "13:00 TIME\n",
      "1 CARDINAL\n",
      "next Thursday DATE\n",
      "1 pm TIME\n",
      "China Paradise Restaurant FAC\n",
      "Dublin GPE\n",
      "the 8th of March DATE\n",
      "China Paradise Restaurant ORG\n",
      "next Friday DATE\n",
      "Petaluma GPE\n",
      "Indian NORP\n",
      "Thai NORP\n",
      "1 CARDINAL\n",
      "Shree Indian Cuisine ORG\n",
      "Petaluma GPE\n",
      "about 11 CARDINAL\n",
      "morning TIME\n",
      "2 CARDINAL\n",
      "Shree Indian Cuisine ORG\n",
      "Petaluma GPE\n",
      "11 am TIME\n",
      "today DATE\n",
      "1 CARDINAL\n",
      "1 CARDINAL\n",
      "Shree Indian Cuisine ORG\n",
      "Petaluma GPE\n",
      "220 Western Avenue FAC\n",
      "a great day DATE\n",
      "Italian NORP\n",
      "Indian NORP\n",
      "San Francisco GPE\n",
      "54 CARDINAL\n",
      "Mint Ristorante Italiano ORG\n",
      "San Francisco GPE\n",
      "quarter past 5 DATE\n",
      "evening TIME\n",
      "54 CARDINAL\n",
      "San Francisco GPE\n",
      "5:15 pm TIME\n",
      "today DATE\n",
      "2 CARDINAL\n",
      "54 CARDINAL\n",
      "2 CARDINAL\n",
      "5:30 pm TIME\n",
      "a good day DATE\n",
      "Foster City GPE\n",
      "Indian NORP\n",
      "One CARDINAL\n",
      "Pastries N Chaat ORG\n",
      "Foster City GPE\n",
      "Tabla Indian Restaurant ORG\n",
      "Foster City GPE\n",
      "this afternoon TIME\n",
      "1:45 TIME\n",
      "One CARDINAL\n",
      "2 CARDINAL\n",
      "Tabla Indian Restaurant ORG\n",
      "Foster City GPE\n",
      "1:45 pm TIME\n",
      "today DATE\n",
      "5:45 pm TIME\n",
      "One CARDINAL\n",
      "5:45 pm TIME\n",
      "next Monday DATE\n",
      "a great day DATE\n",
      "Chinese NORP\n",
      "American NORP\n",
      "Sichuan NORP\n",
      "Mountain View FAC\n",
      "3 CARDINAL\n",
      "One CARDINAL\n",
      "China Wok ORG\n",
      "Mountain View FAC\n",
      "Hunan Chili PERSON\n",
      "Mountain View FAC\n",
      "102 CARDINAL\n",
      "Castro Street FAC\n",
      "4 CARDINAL\n",
      "6 in the evening TIME\n",
      "4 CARDINAL\n",
      "Hunan Chili FAC\n",
      "Mountain View FAC\n",
      "today DATE\n",
      "6 pm TIME\n",
      "650-969-8968 CARDINAL\n",
      "a good day DATE\n",
      "Chinese NORP\n",
      "San Mateo GPE\n",
      "5 CARDINAL\n",
      "Gau Poang PERSON\n",
      "San Mateo GPE\n",
      "SF GPE\n",
      "Alice PERSON\n",
      "San Francisco GPE\n",
      "San Francisco GPE\n",
      "1441 CARDINAL\n",
      "Polk Street FAC\n",
      "12 CARDINAL\n",
      "afternoon TIME\n",
      "2 CARDINAL\n",
      "Bamboo Restaurant FAC\n",
      "San Francisco GPE\n",
      "today DATE\n",
      "12 pm TIME\n",
      "a nice day DATE\n",
      "California GPE\n",
      "SF GPE\n",
      "10 CARDINAL\n",
      "Al's Place ORG\n",
      "San Francisco GPE\n",
      "Alta Msp ORG\n",
      "one CARDINAL\n",
      "San Francisco GPE\n",
      "1275 CARDINAL\n",
      "Minnesota Street FAC\n",
      "San Francisco GPE\n",
      "half past 7 DATE\n",
      "evening TIME\n",
      "San Francisco GPE\n",
      "today DATE\n",
      "7:30 pm TIME\n",
      "2 CARDINAL\n",
      "a nice day DATE\n",
      "American NORP\n",
      "Mexican NORP\n",
      "Foster City GPE\n",
      "2 CARDINAL\n",
      "Tokai NORP\n",
      "Japanese Restaurant GPE\n",
      "Foster GPE\n",
      "Tokyo Village Grill & Sushi ORG\n",
      "Foster City GPE\n",
      "March 4th DATE\n",
      "evening 6:30\n",
      " TIME\n",
      "Tokyo Village Grill & Sushi FAC\n",
      "Foster City GPE\n",
      "2 CARDINAL\n",
      "6:30 pm TIME\n",
      "next monday DATE\n",
      "1221 Chess Drive EVENT\n",
      "650-341 CARDINAL\n",
      "a great Day DATE\n",
      "American NORP\n",
      "Indian NORP\n",
      "San Francisco GPE\n",
      "10 CARDINAL\n",
      "One CARDINAL\n",
      "54 CARDINAL\n",
      "Mint Ristorante Italiano ORG\n",
      "San Francisco GPE\n",
      "415 CARDINAL\n",
      "6 in the evening TIME\n",
      "54 CARDINAL\n",
      "Mint Ristorante Italiano ORG\n",
      "San Francisco GPE\n",
      "6 pm TIME\n",
      "2 CARDINAL\n",
      "today DATE\n",
      "16 CARDINAL\n",
      "Mint Plaza ORG\n",
      "a great day DATE\n",
      "Chinese NORP\n",
      "American NORP\n",
      "Punjabi NORP\n",
      "Pleasanton GPE\n",
      "7 CARDINAL\n",
      "Pleasanton GPE\n",
      "first ORDINAL\n",
      "Aappakadai ORG\n",
      "12:15 DATE\n",
      "2 CARDINAL\n",
      "Aappakadai ORG\n",
      "Pleasanton GPE\n",
      "today DATE\n",
      "12:15 pm TIME\n",
      "925-750-7709 CARDINAL\n",
      "Aappakadai ORG\n",
      "2 CARDINAL\n",
      "12 pm TIME\n",
      "today DATE\n",
      "a good day DATE\n",
      "Sunnyvale GPE\n",
      "3 CARDINAL\n",
      "Gumba ORG\n",
      "Sunnyvale GPE\n",
      "408 CARDINAL\n",
      "176 CARDINAL\n",
      "South Murphy Avenue FAC\n",
      "the evening TIME\n",
      "7:15 TIME\n",
      "Gumba ORG\n",
      "Sunnyvale GPE\n",
      "2 CARDINAL\n",
      "today DATE\n",
      "7:15 pm TIME\n",
      "the afternoon TIME\n",
      "12:30 TIME\n",
      "2 CARDINAL\n",
      "12:30 pm TIME\n",
      "next tuesday DATE\n",
      "one CARDINAL\n",
      "1 CARDINAL\n",
      "next tuesday DATE\n",
      "a good day DATE\n",
      "San Fran GPE\n",
      "10 CARDINAL\n",
      "1760 CARDINAL\n",
      "San Francisco GPE\n",
      "the evening TIME\n",
      "4:45 TIME\n",
      "1760 CARDINAL\n",
      "San Francisco GPE\n",
      "2 CARDINAL\n",
      "today DATE\n",
      "4:45 pm TIME\n",
      "one CARDINAL\n",
      "second ORDINAL\n",
      "6 pm TIME\n",
      "1 CARDINAL\n",
      "6 pm TIME\n",
      "1760 CARDINAL\n",
      "San Jose GPE\n",
      "71 CARDINAL\n",
      "Saint Peter PERSON\n",
      "San Jose GPE\n",
      "San Jose GPE\n",
      "Hukilau ORG\n",
      "408 CARDINAL\n",
      "Hukilau PERSON\n",
      "230 CARDINAL\n",
      "Jackson Street FAC\n",
      "San Jose GPE\n",
      "Rosie Mccann's PERSON\n",
      "Irish Pub & Restaurant ORG\n",
      "a quarter to 12 DATE\n",
      "morning TIME\n",
      "today DATE\n",
      "11:45 am TIME\n",
      "2 CARDINAL\n",
      "Rosie Mccann's PERSON\n",
      "Irish Pub & Restaurant ORG\n",
      "San Jose GPE\n",
      "355 CARDINAL\n",
      "1060 MONEY\n",
      "408 CARDINAL\n",
      "day DATE\n",
      "Oakland GPE\n",
      "4 CARDINAL\n",
      "One CARDINAL\n",
      "Belotti Ristorante E Bottega PERSON\n",
      "Oakland GPE\n",
      "510 CARDINAL\n",
      "Contrasto ORG\n",
      "Oakland GPE\n",
      "morning TIME\n",
      "Contrasto ORG\n",
      "Oakland GPE\n",
      "11:30 am TIME\n",
      "today DATE\n",
      "2 CARDINAL\n",
      "4 CARDINAL\n",
      "Oakland GPE\n",
      "4 CARDINAL\n",
      "388 CARDINAL\n",
      "Grand Avenue FAC\n",
      "a pleasant day DATE\n",
      "Mexican NORP\n",
      "Indian NORP\n",
      "SFO ORG\n",
      "French NORP\n",
      "San Francisco GPE\n",
      "13:15\n",
      " TIME\n",
      "San Francisco GPE\n",
      "1:15 pm TIME\n",
      "2 CARDINAL\n",
      "today DATE\n",
      "Bistro Aix FAC\n",
      "2 CARDINAL\n",
      "today DATE\n",
      "1:30 pm TIME\n",
      "Dublin GPE\n",
      "1 CARDINAL\n",
      "McDonald's ORG\n",
      "Dublin GPE\n",
      "this evening TIME\n",
      "7:15 TIME\n",
      "McDonald's ORG\n",
      "Dublin GPE\n",
      "7:15 pm TIME\n",
      "2 CARDINAL\n",
      "today DATE\n",
      "7 pm TIME\n",
      "McDonald's ORG\n",
      "Dublin GPE\n",
      "7 pm TIME\n",
      "3 CARDINAL\n",
      "14th of March DATE\n",
      "McDonald's ORG\n",
      "Dublin GPE\n",
      "3 CARDINAL\n",
      "March 14th DATE\n",
      "a wonderful day DATE\n",
      "Milpitas GPE\n",
      "Latin American NORP\n",
      "Baja Cactus ORG\n",
      "Milpitas GPE\n",
      "Santa Clara GPE\n",
      "Santa Clara GPE\n",
      "El Amigo Burrito LOC\n",
      "Santa Clara GPE\n",
      "La Paloma FAC\n",
      "afternoon TIME\n",
      "2 CARDINAL\n",
      "La Paloma FAC\n",
      "Santa Clara GPE\n",
      "12:30 pm TIME\n",
      "today DATE\n",
      "this evening TIME\n",
      "2280 CARDINAL\n",
      "a wonderful day DATE\n",
      "San Francisco GPE\n",
      "Indian NORP\n",
      "Italian NORP\n",
      "American NORP\n",
      "1760 CARDINAL\n",
      "San Francisco GPE\n",
      "10 CARDINAL\n",
      "Academy Bar ORG\n",
      "San Francisco GPE\n",
      "415 CARDINAL\n",
      "12:45 pm TIME\n",
      "12:45 pm TIME\n",
      "today DATE\n",
      "Academy Bar ORG\n",
      "San Francisco GPE\n",
      "2 CARDINAL\n",
      "Oakland GPE\n",
      "Mexican NORP\n",
      "Indian NORP\n",
      "7 CARDINAL\n",
      "One CARDINAL\n",
      "Alamar Kitchen & Bar ORG\n",
      "Oakland GPE\n",
      "Judoku Sushi ORG\n",
      "Oakland GPE\n",
      "4 CARDINAL\n",
      "One CARDINAL\n",
      "Lake Chalet FAC\n",
      "Oakland GPE\n",
      "two CARDINAL\n",
      "10:30 in the morning TIME\n",
      "2 CARDINAL\n",
      "Lake Chalet FAC\n",
      "10:30 am TIME\n",
      "today DATE\n",
      "Oakland GPE\n",
      "1 CARDINAL\n",
      "12:30 TIME\n",
      "1 CARDINAL\n",
      "Lake Chalet FAC\n",
      "12:30 pm TIME\n",
      "today DATE\n",
      "Oakland GPE\n",
      "a nice day DATE\n",
      "Alameda GPE\n",
      "Angel Fish ORG\n",
      "Alameda GPE\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "afternoon TIME\n",
      "Angel Fish ORG\n",
      "Alameda GPE\n",
      "1:15 pm TIME\n",
      "today DATE\n",
      "2 CARDINAL\n",
      "today DATE\n",
      "a wonderful day DATE\n",
      "tonight TIME\n",
      "Chinese NORP\n",
      "tonight TIME\n",
      "Millbrae GPE\n",
      "1 CARDINAL\n",
      "Millbrae GPE\n",
      "Tai Wu Restaurant ORG\n",
      "evening 7:30 TIME\n",
      "today DATE\n",
      "Tai Wu Restaurant ORG\n",
      "2 CARDINAL\n",
      "7:30 pm TIME\n",
      "Millbrae GPE\n",
      "300 CARDINAL\n",
      "El Camino Real ORG\n",
      "day DATE\n",
      "Campbell ORG\n",
      "Campbell ORG\n",
      "Effie ORG\n",
      "Restaurant & Bar ORG\n",
      "Effie ORG\n",
      "331 CARDINAL\n",
      "West Hacienda Avenue FAC\n",
      "the 4th of this month DATE\n",
      "the evening TIME\n",
      "6:30 TIME\n",
      "2 CARDINAL\n",
      "Effie ORG\n",
      "Restaurant & Bar ORG\n",
      "Campbell ORG\n",
      "6:30 pm TIME\n",
      "next Monday DATE\n",
      "the 1st DATE\n",
      "Campbell ORG\n",
      "today DATE\n",
      "3 CARDINAL\n",
      "6:30 pm TIME\n",
      "3 CARDINAL\n",
      "408 CARDINAL\n",
      "a great day DATE\n",
      "Mexican NORP\n",
      "American NORP\n",
      "Noodles PERSON\n",
      "Berkeley GPE\n",
      "6 CARDINAL\n",
      "Crave Asia LOC\n",
      "Berkeley GPE\n",
      "510 CARDINAL\n",
      "11:30 in the morning TIME\n",
      "one CARDINAL\n",
      "2 CARDINAL\n",
      "Crave Asia LOC\n",
      "Berkeley GPE\n",
      "11:30 AM TIME\n",
      "today DATE\n",
      "1301 CARDINAL\n",
      "San Pablo Avenue LOC\n",
      "a great day DATE\n",
      "Petaluma GPE\n",
      "4 CARDINAL\n",
      "Petaluma GPE\n",
      "Brixx Pizzeria ORG\n",
      "Cucina Paradiso ORG\n",
      "Petaluma GPE\n",
      "Risibisi ORG\n",
      "Petaluma GPE\n",
      "154 CARDINAL\n",
      "Petaluma Boulevard North FAC\n",
      "half past 1 DATE\n",
      "afternoon TIME\n",
      "Petaluma GPE\n",
      "1:30 pm TIME\n",
      "today DATE\n",
      "2 CARDINAL\n",
      "707 CARDINAL\n",
      "a great day DATE\n",
      "American NORP\n",
      "Mexican NORP\n",
      "San Fran GPE\n",
      "1300 DATE\n",
      "Fillmore GPE\n",
      "San Francisco GPE\n",
      "the Absinthe Brasserie & Bar ORG\n",
      "San Francisco GPE\n",
      "Aperto PERSON\n",
      "San Francisco GPE\n",
      "1434 18th QUANTITY\n",
      "11:00 DATE\n",
      "2 CARDINAL\n",
      "11 am TIME\n",
      "Aperto GPE\n",
      "San Francisco GPE\n",
      "today DATE\n",
      "415 CARDINAL\n",
      "day DATE\n",
      "Pleasanton GPE\n",
      "American NORP\n",
      "Mexican NORP\n",
      "Japanese NORP\n",
      "Osaka GPE\n",
      "4 CARDINAL\n",
      "Japanese NORP\n",
      "Meiko Sushi PERSON\n",
      "Pleasanton GPE\n",
      "2 CARDINAL\n",
      "Pleasanton GPE\n",
      "P.f ORG\n",
      "Chang PERSON\n",
      "7:45 in the evening TIME\n",
      "2 CARDINAL\n",
      "P.F. Chang's PERSON\n",
      "Pleasanton GPE\n",
      "7:45 pm TIME\n",
      "today DATE\n",
      "P.F. Chang's PERSON\n",
      "2 CARDINAL\n",
      "today DATE\n",
      "7:30 pm TIME\n",
      "1330 CARDINAL\n",
      "Stoneridge Mall Road FAC\n",
      "925-224-9916 CARDINAL\n",
      "today DATE\n",
      "Petaluma GPE\n",
      "Italian NORP\n",
      "Petaluma GPE\n",
      "4 CARDINAL\n",
      "Brixx Pizzeria ORG\n",
      "707 CARDINAL\n",
      "six CARDINAL\n",
      "5 pm TIME\n",
      "5 pm TIME\n",
      "today DATE\n",
      "6 CARDINAL\n",
      "Brixx Pizzeria ORG\n",
      "Petaluma GPE\n",
      "16 CARDINAL\n",
      "Kentucky Street FAC\n",
      "Pleasanton GPE\n",
      "American NORP\n",
      "Indian NORP\n",
      "Izakaya food ORG\n",
      "Meiko Sushi LOC\n",
      "Pleasanton GPE\n",
      "12:30 pm TIME\n",
      "2 CARDINAL\n",
      "Meiko Sushi ORG\n",
      "Pleasanton GPE\n",
      "12:30 pm TIME\n",
      "today DATE\n",
      "2 CARDINAL\n",
      "Meiko Sushi ORG\n",
      "Pleasanton GPE\n",
      "7:15 pm TIME\n",
      "today DATE\n",
      "2 CARDINAL\n",
      "Meiko Sushi ORG\n",
      "Pleasanton GPE\n",
      "7:30 pm TIME\n",
      "today DATE\n",
      "4301 Valley Avenue FAC\n",
      "a great day ahead DATE\n",
      "Punjabi NORP\n",
      "SFO ORG\n",
      "10 CARDINAL\n",
      "San Francisco GPE\n",
      "4.2 CARDINAL\n",
      "11 CARDINAL\n",
      "the morning TIME\n",
      "2 CARDINAL\n",
      "11 CARDINAL\n",
      "Amber ORG\n",
      "San Francisco GPE\n",
      "2 CARDINAL\n",
      "today DATE\n",
      "today DATE\n",
      "a great day DATE\n",
      "San Jose GPE\n",
      "Freshwater NORP\n",
      "10 CARDINAL\n",
      "San Jose GPE\n",
      "Ariake PERSON\n",
      "the 10th of this month DATE\n",
      "7 in the evening TIME\n",
      "2 CARDINAL\n",
      "Ariake GPE\n",
      "San Jose GPE\n",
      "March 10th DATE\n",
      "6:45 pm TIME\n",
      "1 CARDINAL\n",
      "afternoon TIME\n",
      "2 CARDINAL\n",
      "Ariake GPE\n",
      "San Jose GPE\n",
      "March 10th DATE\n",
      "1 pm TIME\n",
      "Tuesday next week DATE\n",
      "Ariake GPE\n",
      "next Tuesday DATE\n",
      "a good day DATE\n",
      "San Francisco GPE\n",
      "54 CARDINAL\n",
      "San Francisco GPE\n",
      "5:30 in TIME\n",
      "evening TIME\n",
      "Today DATE\n",
      "5:30 pm TIME\n",
      "San Francisco GPE\n",
      "54 CARDINAL\n",
      "Mint Ristorante Italiano ORG\n",
      "2 CARDINAL\n",
      "March 3rd DATE\n",
      "3 CARDINAL\n",
      "3 CARDINAL\n",
      "tomorrow DATE\n",
      "11:30 am TIME\n",
      "7th of March DATE\n",
      "11:30 am TIME\n",
      "next Thursday DATE\n",
      "a wonderful day DATE\n",
      "Mexican NORP\n",
      "American NORP\n",
      "San Francisco GPE\n",
      "San Francisco GPE\n",
      "Ala Romana LOC\n",
      "Che Fico PERSON\n",
      "San Francisco GPE\n",
      "Mexican NORP\n",
      "3 CARDINAL\n",
      "Mexican NORP\n",
      "Cala GPE\n",
      "San Francisco GPE\n",
      "Cala GPE\n",
      "one CARDINAL\n",
      "March 7th DATE\n",
      "Cala GPE\n",
      "San Francisco GPE\n",
      "1 pm TIME\n",
      "next Thursday DATE\n",
      "2 CARDINAL\n",
      "a quarter past 12 DATE\n",
      "afternoon TIME\n",
      "12:15 pm TIME\n",
      "2 CARDINAL\n",
      "Cala GPE\n",
      "next Thursday DATE\n",
      "12 pm TIME\n",
      "2 CARDINAL\n",
      "a good day DATE\n",
      "Milpitas GPE\n",
      "2 CARDINAL\n",
      "Italian NORP\n",
      "Milpitas GPE\n",
      "408 CARDINAL\n",
      "three CARDINAL\n",
      "1 pm TIME\n",
      "3 CARDINAL\n",
      "Olive Garden Italian Restaurant FAC\n",
      "Milpitas GPE\n",
      "today DATE\n",
      "1 pm TIME\n",
      "4 CARDINAL\n",
      "4 CARDINAL\n",
      "Olive Garden Italian Restaurant FAC\n",
      "Milpitas GPE\n",
      "today DATE\n",
      "1 pm TIME\n",
      "evening TIME\n",
      "Olive Garden Italian Restaurant FAC\n",
      "6:30 pm TIME\n",
      "1350 CARDINAL\n",
      "Great Mall Drive FAC\n",
      "evening TIME\n",
      "Italian NORP\n",
      "Chinese NORP\n",
      "San Jose GPE\n",
      "1 CARDINAL\n",
      "San Jose GPE\n",
      "Mcdonald's ORG\n",
      "777 CARDINAL\n",
      "Road PRODUCT\n",
      "tomorrow DATE\n",
      "11:30 in the morning TIME\n",
      "2 CARDINAL\n",
      "Mcdonald's ORG\n",
      "San Jose GPE\n",
      "11:30 am on day TIME\n",
      "tomorrow DATE\n",
      "one CARDINAL\n",
      "1 CARDINAL\n",
      "San Jose GPE\n",
      "3 CARDINAL\n",
      "2nd of March DATE\n",
      "3 CARDINAL\n",
      "tomorrow DATE\n",
      "a great day DATE\n",
      "Korean Barbeque PERSON\n",
      "Santa Clara GPE\n",
      "1 CARDINAL\n",
      "Restaurant Chungdam ORG\n",
      "Santa Clara GPE\n",
      "20:00 DATE\n",
      "one CARDINAL\n",
      "Restaurant Chungdam ORG\n",
      "Santa Clara GPE\n",
      "8 pm TIME\n",
      "today DATE\n",
      "2 CARDINAL\n",
      "7 in the evening TIME\n",
      "one CARDINAL\n",
      "Restaurant Chungdam ORG\n",
      "Santa Clara GPE\n",
      "7 pm TIME\n",
      "today DATE\n",
      "2 CARDINAL\n",
      "day DATE\n",
      "Mexican NORP\n",
      "Italian NORP\n",
      "SFO ORG\n",
      "Japanese NORP\n",
      "4 CARDINAL\n",
      "Izakaya Sushi Ran ORG\n",
      "San Francisco GPE\n",
      "415 CARDINAL\n",
      "evening 5:45 TIME\n",
      "2 CARDINAL\n",
      "Izakaya Sushi Ran ORG\n",
      "San Francisco GPE\n",
      "today DATE\n",
      "5:45 pm TIME\n",
      "1 CARDINAL\n",
      "8th ORDINAL\n",
      "1 CARDINAL\n",
      "next Friday DATE\n",
      "3 CARDINAL\n",
      "11:15 DATE\n",
      "3 CARDINAL\n",
      "11:15 am TIME\n",
      "a great day DATE\n",
      "American NORP\n",
      "Chinese NORP\n",
      "Berkeley GPE\n",
      "Berkeley GPE\n",
      "2 CARDINAL\n",
      "B&b Kitchen & Wine Bar ORG\n",
      "today DATE\n",
      "1:30 pm TIME\n",
      "Berkeley GPE\n",
      "seven pm TIME\n",
      "2 CARDINAL\n",
      "B&b Kitchen & Wine Bar ORG\n",
      "today DATE\n",
      "7 pm TIME\n",
      "Berkeley GPE\n",
      "a great day DATE\n",
      "American NORP\n",
      "Indian NORP\n",
      "Mexican NORP\n",
      "San Mateo GPE\n",
      "San Mateo GPE\n",
      "Celia PERSON\n",
      "11:30 am TIME\n",
      "2 CARDINAL\n",
      "Celia PERSON\n",
      "San Mateo GPE\n",
      "today DATE\n",
      "11:30 am TIME\n",
      "650-349-0165 CARDINAL\n",
      "SFO ORG\n",
      "10 CARDINAL\n",
      "One CARDINAL\n",
      "Baker Street Bistro FAC\n",
      "San Francisco GPE\n",
      "San Francisco GPE\n",
      "evening 5:30 TIME\n",
      "Blue Mermaid PERSON\n",
      "San Francisco GPE\n",
      "5:30 pm TIME\n",
      "2 CARDINAL\n",
      "today DATE\n",
      "471 Jefferson Street FAC\n",
      "a wonderful day DATE\n",
      "SFO ORG\n",
      "Oakland GPE\n",
      "Oakland GPE\n",
      "Dyafa ORG\n",
      "1 CARDINAL\n",
      "Marhaba Indian & Pakistani Halal Cuisine ORG\n",
      "Oakland GPE\n",
      "1437 Franklin Street FAC\n",
      "13:30 TIME\n",
      "Marhaba Indian & Pakistani Halal Cuisine ORG\n",
      "Oakland GPE\n",
      "2 CARDINAL\n",
      "today DATE\n",
      "1:30 pm TIME\n",
      "1 CARDINAL\n",
      "1 CARDINAL\n",
      "today DATE\n",
      "a nice day DATE\n",
      "Calistoga GPE\n",
      "Taiwanese NORP\n",
      "Calistoga GPE\n",
      "1354 CARDINAL\n",
      "29 CARDINAL\n",
      "4 CARDINAL\n",
      "1:30 in the afternoon TIME\n",
      "4 CARDINAL\n",
      "Soo Yuan Restaurant ORG\n",
      "Calistoga GPE\n",
      "1:30 pm TIME\n",
      "today DATE\n",
      "a great day DATE\n",
      "Berkeley GPE\n",
      "Breakfast PRODUCT\n",
      "1 CARDINAL\n",
      "the Venus Restaurant ORG\n",
      "Berkeley GPE\n",
      "2327 Shattuck Avenue FAC\n",
      "1 CARDINAL\n",
      "12:15 in the afternoon TIME\n",
      "1 CARDINAL\n",
      "12:15 pm TIME\n",
      "today DATE\n",
      "the Venus Restaurant FAC\n",
      "Berkeley GPE\n",
      "evening 6:45 TIME\n",
      "1 CARDINAL\n",
      "6:45 pm TIME\n",
      "510-540-5950 CARDINAL\n",
      "1 CARDINAL\n",
      "today DATE\n",
      "7 pm TIME\n",
      "the Venus Restaurant FAC\n",
      "today DATE\n",
      "a great day DATE\n",
      "Sushi GPE\n",
      "San Rafael GPE\n",
      "Sushi To Dai For ORG\n",
      "San Rafael GPE\n",
      "816 4th Street FAC\n",
      "415 CARDINAL\n",
      "Sichuan NORP\n",
      "Sichuan NORP\n",
      "San Rafael GPE\n",
      "7 in the evening TIME\n",
      "Monday next week DATE\n",
      "Yet Wah Restaurant FAC\n",
      "San Rafael GPE\n",
      "7 pm TIME\n",
      "next Monday DATE\n",
      "2 CARDINAL\n",
      "day DATE\n",
      "Cotati GPE\n",
      "Chinese NORP\n",
      "American NORP\n",
      "Mi Pueblo FAC\n",
      "Cotati GPE\n",
      "Cotati GPE\n",
      "1 CARDINAL\n",
      "Cotati GPE\n",
      "Mi Ranchito ORG\n",
      "7600 CARDINAL\n",
      "Commerce Boulevard FAC\n",
      "two CARDINAL\n",
      "tonight TIME\n",
      "half past 6 CARDINAL\n",
      "evening TIME\n",
      "2 CARDINAL\n",
      "today DATE\n",
      "6:30 pm TIME\n",
      "Mi Ranchito FAC\n",
      "Cotati GPE\n",
      "707 CARDINAL\n",
      "day DATE\n",
      "San Francisco GPE\n",
      "10 CARDINAL\n",
      "54 CARDINAL\n",
      "Mint Ristorante Italiano ORG\n",
      "San Francisco GPE\n",
      "Sunday this week DATE\n",
      "4 CARDINAL\n",
      "17:30 TIME\n",
      "54 CARDINAL\n",
      "San Francisco GPE\n",
      "5:30 pm TIME\n",
      "the day after DATE\n",
      "tomorrow DATE\n",
      "4 CARDINAL\n",
      "Thursday next week DATE\n",
      "next Thursday DATE\n",
      "4 CARDINAL\n",
      "19:30 TIME\n",
      "five CARDINAL\n",
      "7:30 pm TIME\n",
      "5 CARDINAL\n",
      "16 CARDINAL\n",
      "a good day DATE\n",
      "SF GPE\n",
      "Thai NORP\n",
      "10 CARDINAL\n",
      "San Francisco GPE\n",
      "one CARDINAL\n",
      "Baan Thai House & Wine Bar ORG\n",
      "415 CARDINAL\n",
      "534 CARDINAL\n",
      "Irving Street FAC\n",
      "6 in the evening TIME\n",
      "6 pm TIME\n",
      "today DATE\n",
      "2 CARDINAL\n",
      "Baan Thai House & Wine Bar ORG\n",
      "San Francisco GPE\n",
      "one CARDINAL\n",
      "6 pm TIME\n",
      "1 CARDINAL\n",
      "American NORP\n",
      "Italian NORP\n",
      "San Bruno GPE\n",
      "Taiwanese NORP\n",
      "China Garden FAC\n",
      "San Bruno GPE\n",
      "586 CARDINAL\n",
      "San Mateo Avenue GPE\n",
      "Dumpling Era ORG\n",
      "San Bruno GPE\n",
      "1 CARDINAL\n",
      "Ming's ORG\n",
      "San Bruno GPE\n",
      "half past 12 CARDINAL\n",
      "6 CARDINAL\n",
      "Ming's ORG\n",
      "San Bruno GPE\n",
      "6 CARDINAL\n",
      "today DATE\n",
      "12:30 pm TIME\n",
      "eleven am TIME\n",
      "Ming's ORG\n",
      "San Bruno GPE\n",
      "11 am TIME\n",
      "today DATE\n",
      "6 CARDINAL\n",
      "four CARDINAL\n",
      "11 am TIME\n",
      "4 CARDINAL\n",
      "day DATE\n",
      "Greek NORP\n",
      "San Carlos GPE\n",
      "2 CARDINAL\n",
      "Greek NORP\n",
      "Agora PERSON\n",
      "San Carlos GPE\n",
      "One CARDINAL\n",
      "Spicy Noodles PERSON\n",
      "2 CARDINAL\n",
      "Asya Restaurant ORG\n",
      "San Carlos GPE\n",
      "1 CARDINAL\n",
      "Thai Time ORG\n",
      "San Carlos GPE\n",
      "12:45 DATE\n",
      "2 CARDINAL\n",
      "Thai Time EVENT\n",
      "San Carlos GPE\n",
      "12:45 pm TIME\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "650-596-8433 CARDINAL\n",
      "1240 CARDINAL\n",
      "a good day DATE\n",
      "Petaluma GPE\n",
      "Indian NORP\n",
      "Chinese NORP\n",
      "Asian Fusion ORG\n",
      "Flowering Tea House ORG\n",
      "Petaluma GPE\n",
      "5:15 in the evening TIME\n",
      "next Friday DATE\n",
      "2 CARDINAL\n",
      "next Friday DATE\n",
      "5:15 PM TIME\n",
      "Petaluma GPE\n",
      "the Flowering Tea House FAC\n",
      "1 CARDINAL\n",
      "the time quarter past 1 DATE\n",
      "afternoon TIME\n",
      "1 CARDINAL\n",
      "1:15 pm TIME\n",
      "the 13th of this month DATE\n",
      "3 CARDINAL\n",
      "March 13th DATE\n",
      "3 CARDINAL\n",
      "Flowering Tea House ORG\n",
      "3 CARDINAL\n",
      "March 13th DATE\n",
      "1 pm TIME\n",
      "a nice day DATE\n",
      "Chinese NORP\n",
      "American NORP\n",
      "Campbell PERSON\n",
      "5 CARDINAL\n",
      "Bellagio Italian Restaurant ORG\n",
      "Campbell ORG\n",
      "11:30 TIME\n",
      "11:30 am TIME\n",
      "today DATE\n",
      "2 CARDINAL\n",
      "A Bellagio Italian Restaurant FAC\n",
      "Campbell ORG\n",
      "408 CARDINAL\n",
      "33 CARDINAL\n",
      "South Central Avenue LOC\n",
      "San Francisco GPE\n",
      "Burgers NORP\n",
      "8 CARDINAL\n",
      "5a5 CARDINAL\n",
      "Steak Lounge ORG\n",
      "San Francisco GPE\n",
      "Alexander's Steakhouse ORG\n",
      "San Francisco GPE\n",
      "San Francisco GPE\n",
      "Bourbon Steak Restaurant WORK_OF_ART\n",
      "one CARDINAL\n",
      "13:30 CARDINAL\n",
      "Bourbon Steak Restaurant FAC\n",
      "San Francisco GPE\n",
      "1:30 pm TIME\n",
      "today DATE\n",
      "2 CARDINAL\n",
      "a great day DATE\n",
      "Pleasanton GPE\n",
      "American NORP\n",
      "Indian NORP\n",
      "Barbecue LOC\n",
      "2 CARDINAL\n",
      "Baci Bistro FAC\n",
      "Pleasanton GPE\n",
      "925-600-0600 CARDINAL\n",
      "13:45 TIME\n",
      "2 CARDINAL\n",
      "Baci Bistro FAC\n",
      "Pleasanton GPE\n",
      "1:45 pm TIME\n",
      "today DATE\n",
      "500 Main Street FAC\n",
      "2 CARDINAL\n",
      "Baci Bistro FAC\n",
      "1:30 pm TIME\n",
      "today DATE\n",
      "day DATE\n",
      "American NORP\n",
      "Indian NORP\n",
      "Sushi NORP\n",
      "San Francisco GPE\n",
      "San Francisco GPE\n",
      "2 CARDINAL\n",
      "Japanese NORP\n",
      "one CARDINAL\n",
      "afternoon 12 DATE\n",
      "1 CARDINAL\n",
      "today DATE\n",
      "12 pm TIME\n",
      "2 CARDINAL\n",
      "Japanese NORP\n",
      "San Francisco GPE\n",
      "three CARDINAL\n",
      "3 CARDINAL\n",
      "today DATE\n",
      "12 pm TIME\n",
      "2 CARDINAL\n",
      "Japanese NORP\n",
      "San Francisco GPE\n",
      "second ORDINAL\n",
      "12:30 in TIME\n",
      "afternoon TIME\n",
      "3 CARDINAL\n",
      "today DATE\n",
      "12:30 pm TIME\n",
      "2 CARDINAL\n",
      "Japanese NORP\n",
      "San Francisco GPE\n",
      "601 CARDINAL\n",
      "Van Ness Avenue ORG\n",
      "415 CARDINAL\n",
      "a good day DATE\n",
      "Antioch ORG\n",
      "American NORP\n",
      "Mexican NORP\n",
      "Antioch GPE\n",
      "Okawa Japanese Restaurant ORG\n",
      "one CARDINAL\n",
      "4:30 pm TIME\n",
      "Okawa Japanese Restaurant ORG\n",
      "today DATE\n",
      "Antioch ORG\n",
      "1 CARDINAL\n",
      "4:30 pm TIME\n",
      "18:00 DATE\n",
      "Okawa Japanese Restaurant ORG\n",
      "Antioch GPE\n",
      "6 pm TIME\n",
      "today DATE\n",
      "1 CARDINAL\n",
      "a great day DATE\n",
      "Italian NORP\n",
      "Indian NORP\n",
      "Vallejo GPE\n",
      "Mcdonald's ORG\n",
      "Vallejo GPE\n",
      "1 CARDINAL\n",
      "afternoon TIME\n",
      "Mcdonald's ORG\n",
      "Vallejo GPE\n",
      "2 CARDINAL\n",
      "1 pm TIME\n",
      "today DATE\n",
      "Mountain View FAC\n",
      "French NORP\n",
      "La Fontaine PERSON\n",
      "Mountain View FAC\n",
      "650 CARDINAL\n",
      "186 CARDINAL\n",
      "Castro Street FAC\n",
      "SF GPE\n",
      "10 CARDINAL\n",
      "Amelie San Francisco GPE\n",
      "San Francisco GPE\n",
      "morning 10:30 TIME\n",
      "2 CARDINAL\n",
      "Amelie San Francisco GPE\n",
      "10:30 am TIME\n",
      "today DATE\n",
      "San Francisco GPE\n",
      "eleven in the morning TIME\n",
      "2 CARDINAL\n",
      "Amelie San Francisco GPE\n",
      "11 am today TIME\n",
      "San Francisco GPE\n",
      "afternoon TIME\n",
      "San Francisco GPE\n",
      "Russian NORP\n",
      "1 CARDINAL\n",
      "Red Tavern ORG\n",
      "San Francisco GPE\n",
      "2229 CARDINAL\n",
      "next Thursday DATE\n",
      "6:15 in the evening TIME\n",
      "2 CARDINAL\n",
      "Red Tavern ORG\n",
      "San Francisco GPE\n",
      "next Thursday DATE\n",
      "6:15 pm TIME\n",
      "415 CARDINAL\n",
      "a good day DATE\n",
      "American NORP\n",
      "Chinese NORP\n",
      "San francisco GPE\n",
      "1760 CARDINAL\n",
      "San francisco GPE\n",
      "1760 DATE\n",
      "4:45 TIME\n",
      "1760 DATE\n",
      "San francisco GPE\n",
      "4:45 pm TIME\n",
      "today DATE\n",
      "2 CARDINAL\n",
      "a great day DATE\n",
      "Santa Rosa GPE\n",
      "4 CARDINAL\n",
      "Santa Rosa GPE\n",
      "707 CARDINAL\n",
      "Santa Rosa GPE\n",
      "10:45 in the morning TIME\n",
      "2 CARDINAL\n",
      "Santa Rosa GPE\n",
      "10:45 am TIME\n",
      "next Monday DATE\n",
      "2446 CARDINAL\n",
      "2 CARDINAL\n",
      "next Monday DATE\n",
      "11 am TIME\n",
      "a great day DATE\n",
      "asian NORP\n",
      "SFO ORG\n",
      "10 CARDINAL\n",
      "San Francisco GPE\n",
      "8 CARDINAL\n",
      "San Francisco GPE\n",
      "Butterfly Restaurant ORG\n",
      "33 CARDINAL\n",
      "Embarcadero GPE\n",
      "San Francisco GPE\n",
      "Chaat Corner FAC\n",
      "evening TIME\n",
      "2 CARDINAL\n",
      "Chaat Corner FAC\n",
      "San Francisco GPE\n",
      "today DATE\n",
      "7:30 pm TIME\n",
      "March 14th DATE\n",
      "San Francisco GPE\n",
      "March 14th DATE\n",
      "three CARDINAL\n",
      "3 CARDINAL\n",
      "March 14th DATE\n",
      "Mediterranean LOC\n",
      "Redwood City GPE\n",
      "only 1 CARDINAL\n",
      "Redwood City GPE\n",
      "17:30 TIME\n",
      "Redwood City GPE\n",
      "5:30 pm TIME\n",
      "today DATE\n",
      "2 CARDINAL\n",
      "Chinese NORP\n",
      "Mexican NORP\n",
      "Italian NORP\n",
      "American NORP\n",
      "American NORP\n",
      "Fairfield GPE\n",
      "American NORP\n",
      "Fairfield GPE\n",
      "Chuck E. Cheese's PERSON\n",
      "American NORP\n",
      "Red Lobster PERSON\n",
      "American NORP\n",
      "Fairfield GPE\n",
      "Texas GPE\n",
      "Roadhouse GPE\n",
      "American NORP\n",
      "Fairfield GPE\n",
      "3333 CARDINAL\n",
      "the morning TIME\n",
      "11:15 am TIME\n",
      "Texas GPE\n",
      "Roadhouse GPE\n",
      "Fairfield GPE\n",
      "2 CARDINAL\n",
      "today DATE\n",
      "11:15 am TIME\n",
      "13th of this month DATE\n",
      "Texas Roadhouse ORG\n",
      "March 13th DATE\n",
      "707 CARDINAL\n",
      "a wonderful day DATE\n",
      "Oakland GPE\n",
      "Italian NORP\n",
      "American NORP\n",
      "Gastrobar ORG\n",
      "1 CARDINAL\n",
      "Burger Oakland ORG\n",
      "Oakland GPE\n",
      "2100 CARDINAL\n",
      "Franklin Street FAC\n",
      "the 11th of March DATE\n",
      "6 CARDINAL\n",
      "two CARDINAL\n",
      "Umami Burger Oakland ORG\n",
      "Oakland GPE\n",
      "6 pm TIME\n",
      "March 11th DATE\n",
      "2 CARDINAL\n",
      "510 CARDINAL\n",
      "a good day DATE\n",
      "Oakland GPE\n",
      "American NORP\n",
      "Mexican NORP\n",
      "Alamar Kitchen & Bar ORG\n",
      "Oakland GPE\n",
      "510 CARDINAL\n",
      "6:15 pm TIME\n",
      "6:15 pm TIME\n",
      "2 CARDINAL\n",
      "today DATE\n",
      "Alamar Kitchen & Bar ORG\n",
      "Oakland GPE\n",
      "evening TIME\n",
      "5:30 TIME\n",
      "2 CARDINAL\n",
      "5:30 pm TIME\n",
      "today DATE\n",
      "Alamar Kitchen & Bar ORG\n",
      "Oakland GPE\n",
      "100 CARDINAL\n",
      "Grand Avenue FAC\n",
      "111 MONEY\n",
      "a nice day DATE\n",
      "Santa Clara GPE\n",
      "Southern NORP\n",
      "4 CARDINAL\n",
      "Bennigan PERSON\n",
      "Santa Clara GPE\n",
      "two CARDINAL\n",
      "18:45 DATE\n",
      "2 CARDINAL\n",
      "Bennigan PERSON\n",
      "Santa Clara GPE\n",
      "6:45 pm TIME\n",
      "today DATE\n",
      "11:30 am TIME\n",
      "8th ORDINAL\n",
      "11:30 am TIME\n",
      "next Friday DATE\n",
      "4150 Great America Parkway FAC\n",
      "Burlingame GPE\n",
      "Mexican NORP\n",
      "Indian NORP\n",
      "4 CARDINAL\n",
      "Burlingame GPE\n",
      "1:15 pm TIME\n",
      "Burlingame GPE\n",
      "37 North PERCENT\n",
      "today DATE\n",
      "1:15 pm TIME\n",
      "2 CARDINAL\n",
      "650-373-2237 CARDINAL\n",
      "Burlingame GPE\n",
      "37 North PERCENT\n",
      "today DATE\n",
      "1:30 pm TIME\n",
      "2 CARDINAL\n",
      "a great day DATE\n",
      "Italian NORP\n",
      "Pleasanton GPE\n",
      "De La Torre's PERSON\n",
      "Pleasanton GPE\n",
      "Forno Vecchio ORG\n",
      "Pleasanton GPE\n",
      "925 CARDINAL\n",
      "Gay 90's ORG\n",
      "Pizza Co. ORG\n",
      "Pleasanton GPE\n",
      "10th of this month DATE\n",
      "Gay 90's ORG\n",
      "Pizza Co. ORG\n",
      "2 CARDINAL\n",
      "Pleasanton GPE\n",
      "1:15 pm TIME\n",
      "March 10th DATE\n",
      "288 Main Street FAC\n",
      "a wonderful day DATE\n",
      "Taiwanese NORP\n",
      "SFO ORG\n",
      "Alice PERSON\n",
      "San Francisco GPE\n",
      "9 CARDINAL\n",
      "San Francisco GPE\n",
      "415 CARDINAL\n",
      "1441 CARDINAL\n",
      "Polk Street FAC\n",
      "Bei Fang PERSON\n",
      "San Francisco GPE\n",
      "three CARDINAL\n",
      "afternoon TIME\n",
      "San Francisco GPE\n",
      "3 CARDINAL\n",
      "today DATE\n",
      "1:30 pm TIME\n",
      "a wonderful day DATE\n",
      "Oakland GPE\n",
      "African NORP\n",
      "2 CARDINAL\n",
      "Cafe Eritrea ORG\n",
      "Oakland GPE\n",
      "1 CARDINAL\n",
      "Miliki Restaurant ORG\n",
      "Oakland GPE\n",
      "3725 MacArthur Boulevard FAC\n",
      "the 12th DATE\n",
      "11:15 in the morning TIME\n",
      "Miliki Restaurant ORG\n",
      "Oakland GPE\n",
      "11:15 am TIME\n",
      "2 CARDINAL\n",
      "March 12th DATE\n",
      "the 3rd of this month DATE\n",
      "2 CARDINAL\n",
      "Miliki Restaurant ORG\n",
      "Oakland GPE\n",
      "11:15 am TIME\n",
      "the day after DATE\n",
      "tomorrow DATE\n",
      "12:30 TIME\n",
      "12:30 pm TIME\n",
      "2 CARDINAL\n",
      "today DATE\n",
      "a great day DATE\n",
      "Walnut Creek GPE\n",
      "3 CARDINAL\n",
      "Walnut Creek GPE\n",
      "Panera Bread ORG\n",
      "1140 CARDINAL\n",
      "6 CARDINAL\n",
      "2 CARDINAL\n",
      "Panera Bread ORG\n",
      "Walnut Creek GPE\n",
      "6 pm TIME\n",
      "today DATE\n",
      "San Mateo GPE\n",
      "Indian NORP\n",
      "Italian NORP\n",
      "Central Park Bistro FAC\n",
      "San Mateo GPE\n",
      "181 CARDINAL\n",
      "East 4th Avenue ORG\n",
      "5:45 in the evening TIME\n",
      "Central Park Bistro FAC\n",
      "San Mateo GPE\n",
      "5:45 pm TIME\n",
      "2 CARDINAL\n",
      "today DATE\n",
      "Central Park Bistro FAC\n",
      "2 CARDINAL\n",
      "today DATE\n",
      "6 pm\n",
      " TIME\n",
      "one CARDINAL\n",
      "Indian NORP\n",
      "Newark GPE\n",
      "4 CARDINAL\n",
      "Biryani Pot PERSON\n",
      "Newark GPE\n",
      "7:30 in the evening TIME\n",
      "2 CARDINAL\n",
      "today DATE\n",
      "7:30 pm TIME\n",
      "the Biryani Pot FAC\n",
      "Newark GPE\n",
      "day DATE\n",
      "Italian NORP\n",
      "Mexican NORP\n",
      "San Bruno GPE\n",
      "Latin American NORP\n",
      "Celia PERSON\n",
      "Mexican Restaurant ORG\n",
      "San Bruno GPE\n",
      "1 CARDINAL\n",
      "Chili's Grill & Bar ORG\n",
      "San Bruno GPE\n",
      "Next Friday DATE\n",
      "six CARDINAL\n",
      "9th ORDINAL\n",
      "5:30 in the evening TIME\n",
      "Chili's Grill & Bar ORG\n",
      "San Bruno GPE\n",
      "5:30 pm TIME\n",
      "March 9th DATE\n",
      "6 CARDINAL\n",
      "6:30 in the evening TIME\n",
      "San Bruno GPE\n",
      "6:30 pm TIME\n",
      "2 CARDINAL\n",
      "Chili's Grill & Bar ORG\n",
      "San Bruno GPE\n",
      "6:30 pm TIME\n",
      "March 9th DATE\n",
      "2 CARDINAL\n",
      "a good day DATE\n",
      "Mexican NORP\n",
      "Chinese NORP\n",
      "Thai NORP\n",
      "Oriental NORP\n",
      "Berkeley GPE\n",
      "Mount Everest Restaurant ORG\n",
      "Berkeley GPE\n",
      "1 CARDINAL\n",
      "1 CARDINAL\n",
      "the 13th DATE\n",
      "12:30 TIME\n",
      "Mount Everest Restaurant FAC\n",
      "Berkeley GPE\n",
      "March 13th DATE\n",
      "12:30 pm TIME\n",
      "1 CARDINAL\n",
      "Mount Everest Restaurant FAC\n",
      "2598 Telegraph Avenue ORG\n",
      "510 CARDINAL\n",
      "a nice day DATE\n",
      "SF GPE\n",
      "Chinese NORP\n",
      "Indian NORP\n",
      "10 CARDINAL\n",
      "Alice PERSON\n",
      "San Francisco GPE\n",
      "415 CARDINAL\n",
      "1599 CARDINAL\n",
      "Sanchez Street FAC\n",
      "the 1st of this month DATE\n",
      "around 11:45 TIME\n",
      "2 CARDINAL\n",
      "Alice PERSON\n",
      "San Francisco GPE\n",
      "11:45 am TIME\n",
      "today DATE\n",
      "Alice ORG\n",
      "2 CARDINAL\n",
      "today DATE\n",
      "12 pm TIME\n",
      "a great day DATE\n",
      "Morgan Hill GPE\n",
      "Hanami Sushi PERSON\n",
      "Morgan Hill GPE\n",
      "Nara GPE\n",
      "Morgan Hill GPE\n",
      "two CARDINAL\n",
      "afternoon TIME\n",
      "12:30 TIME\n",
      "2 CARDINAL\n",
      "Sushi Nara FAC\n",
      "Morgan Hill LOC\n",
      "12:30 pm TIME\n",
      "today DATE\n",
      "three CARDINAL\n",
      "the 12th DATE\n",
      "3 on March 12th DATE\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "indian NORP\n",
      "chinese NORP\n",
      "2 CARDINAL\n",
      "next wednesday DATE\n",
      "1 pm TIME\n",
      "1 pm TIME\n",
      "next wednesday DATE\n",
      "2 CARDINAL\n",
      "19:00 TIME\n",
      "march 9th DATE\n",
      "7 pm march 9th TIME\n",
      "650 CARDINAL\n",
      "a great day DATE\n",
      "Fremont GPE\n",
      "4 CARDINAL\n",
      "Fremont GPE\n",
      "Aniki ORG\n",
      "510 CARDINAL\n",
      "3810 CARDINAL\n",
      "Mowry Avenue FAC\n",
      "one CARDINAL\n",
      "1 CARDINAL\n",
      "12:30 pm TIME\n",
      "Aniki's Sushi ORG\n",
      "Fremont GPE\n",
      "three CARDINAL\n",
      "the 14th of March DATE\n",
      "3 CARDINAL\n",
      "March 14th DATE\n",
      "1 pm TIME\n",
      "9th ORDINAL\n",
      "1 pm TIME\n",
      "March 9th DATE\n",
      "a good day DATE\n",
      "Asian NORP\n",
      "San Mateo GPE\n",
      "San Mateo GPE\n",
      "Izakaya Ginji GPE\n",
      "1:30 pm TIME\n",
      "1 CARDINAL\n",
      "Izakaya Ginji GPE\n",
      "San Mateo GPE\n",
      "1:30 pm TIME\n",
      "today DATE\n",
      "1 CARDINAL\n",
      "925 CARDINAL\n",
      "today DATE\n",
      "a nice day DATE\n",
      "South San Francisco GPE\n",
      "Italian NORP\n",
      "Mexican NORP\n",
      "3 CARDINAL\n",
      "South San Francisco GPE\n",
      "San Mateo GPE\n",
      "7 CARDINAL\n",
      "Daiki Sushi PERSON\n",
      "San Mateo GPE\n",
      "Dash Japanese Tapas And Sushi ORG\n",
      "San Mateo GPE\n",
      "2 CARDINAL\n",
      "12:15 DATE\n",
      "2 CARDINAL\n",
      "Dash Japanese Tapas And Sushi ORG\n",
      "San Mateo GPE\n",
      "12:15 pm TIME\n",
      "today DATE\n",
      "1 CARDINAL\n",
      "10:45 TIME\n",
      "1 CARDINAL\n",
      "10:45 am TIME\n",
      "Dash Japanese Tapas And Sushi ORG\n",
      "1 CARDINAL\n",
      "today DATE\n",
      "11 am TIME\n",
      "a good day DATE\n",
      "SFO ORG\n",
      "Mexican NORP\n",
      "Italian NORP\n",
      "Pakistani NORP\n",
      "4 CARDINAL\n",
      "Pakistani NORP\n",
      "San Francisco GPE\n",
      "Aslam ORG\n",
      "Rasoi NORP\n",
      "March 14th DATE\n",
      "17:30 TIME\n",
      "Aslam ORG\n",
      "Rasoi NORP\n",
      "San Francisco GPE\n",
      "March 14th DATE\n",
      "5:30 pm TIME\n",
      "2 CARDINAL\n",
      "three CARDINAL\n",
      "Aslam ORG\n",
      "Rasoi NORP\n",
      "San Francisco GPE\n",
      "3 CARDINAL\n",
      "Aslam ORG\n",
      "Rasoi NORP\n",
      "San Francisco GPE\n",
      "415 CARDINAL\n",
      "a good day DATE\n",
      "San Jose GPE\n",
      "4 CARDINAL\n",
      "San Jose GPE\n",
      "4th St Pizza Co ORG\n",
      "San Jose GPE\n",
      "Mimi's Cafe ORG\n",
      "5 CARDINAL\n",
      "the 7th of March DATE\n",
      "four CARDINAL\n",
      "6 pm\n",
      " TIME\n",
      "4 CARDINAL\n",
      "Mimi's Cafe ORG\n",
      "San Jose GPE\n",
      "6 pm TIME\n",
      "next Thursday DATE\n",
      "the 12th of this month DATE\n",
      "twelve pm TIME\n",
      "March 12th DATE\n",
      "12 pm TIME\n",
      "a great day DATE\n",
      "Parisian NORP\n",
      "San Francisco GPE\n",
      "10 CARDINAL\n",
      "One CARDINAL\n",
      "Amelie San Francisco GPE\n",
      "San Francisco GPE\n",
      "about quarter past 11 DATE\n",
      "morning TIME\n",
      "2 CARDINAL\n",
      "11:15 am TIME\n",
      "today DATE\n",
      "Amelie San Francisco GPE\n",
      "San Francisco GPE\n",
      "6 CARDINAL\n",
      "2 CARDINAL\n",
      "6 pm TIME\n",
      "today DATE\n",
      "Amelie San Francisco GPE\n",
      "San Francisco GPE\n",
      "day DATE\n",
      "Indian NORP\n",
      "Italian NORP\n",
      "Pleasanton GPE\n",
      "Taiwanese NORP\n",
      "Pleasanton GPE\n",
      "China Lounge Restaurant & Bar ORG\n",
      "Pleasanton GPE\n",
      "Hunan Chef Chinese Restaurant ORG\n",
      "Pleasanton GPE\n",
      "4285 Valley Avenue ORG\n",
      "12 in the afternoon TIME\n",
      "2 CARDINAL\n",
      "Hunan Chef Chinese Restaurant ORG\n",
      "Pleasanton GPE\n",
      "12 pm TIME\n",
      "today DATE\n",
      "one CARDINAL\n",
      "6th ORDINAL\n",
      "1 CARDINAL\n",
      "next Wednesday DATE\n",
      "three CARDINAL\n",
      "5:15 pm TIME\n",
      "3 CARDINAL\n",
      "5:15 pm TIME\n",
      "a good day DATE\n",
      "Izakaya GPE\n",
      "Gastrobar PERSON\n",
      "St. Helena GPE\n",
      "St. Helena ORG\n",
      "Cook Tavern & Pizzeria ORG\n",
      "1304 CARDINAL\n",
      "2 CARDINAL\n",
      "the Cook Tavern & Pizzeria ORG\n",
      "St. Helena GPE\n",
      "12:30 pm TIME\n",
      "today DATE\n",
      "4 CARDINAL\n",
      "17:00 TIME\n",
      "4 CARDINAL\n",
      "5 pm TIME\n",
      "the 12th of March DATE\n",
      "6 CARDINAL\n",
      "6 CARDINAL\n",
      "March 12th DATE\n",
      "a nice day DATE\n",
      "American NORP\n",
      "Italian NORP\n",
      "Burger ORG\n",
      "San Francisco GPE\n",
      "about 10 CARDINAL\n",
      "San Francisco GPE\n",
      "1760 CARDINAL\n",
      "Tuesday next week DATE\n",
      "evening 5 DATE\n",
      "2 CARDINAL\n",
      "next Tuesday DATE\n",
      "5 pm TIME\n",
      "1760 CARDINAL\n",
      "San Francisco GPE\n",
      "a good day DATE\n",
      "Dublin GPE\n",
      "Chinese NORP\n",
      "7 CARDINAL\n",
      "one CARDINAL\n",
      "China Paradise Restaurant ORG\n",
      "Dublin GPE\n",
      "4580 Dublin Boulevard FAC\n",
      "7th ORDINAL\n",
      "four CARDINAL\n",
      "14:00 DATE\n",
      "4 CARDINAL\n",
      "China Paradise Restaurant FAC\n",
      "Dublin GPE\n",
      "next Thursday DATE\n",
      "2 pm TIME\n",
      "2 CARDINAL\n",
      "2 CARDINAL\n",
      "China Paradise Restaurant FAC\n",
      "Dublin GPE\n",
      "next Thursday DATE\n",
      "2 pm TIME\n",
      "March 9th DATE\n",
      "2 CARDINAL\n",
      "China Paradise Restaurant FAC\n",
      "Dublin GPE\n",
      "March 9th DATE\n",
      "2 pm TIME\n",
      "1:45 DATE\n",
      "2 CARDINAL\n",
      "China Paradise Restaurant FAC\n",
      "Dublin GPE\n",
      "March 9th DATE\n",
      "1:45 pm TIME\n",
      "day DATE\n",
      "one CARDINAL\n",
      "Italian NORP\n",
      "chinese NORP\n",
      "Asian NORP\n",
      "Pleasant Hill GPE\n",
      "China Chef Restaurant ORG\n",
      "Asian NORP\n",
      "Pleasant Hill GPE\n",
      "925 CARDINAL\n",
      "1200 CARDINAL\n",
      "Contra NORP\n",
      "K. MONEY\n",
      "Tuesday next week DATE\n",
      "18:45 TIME\n",
      "2 CARDINAL\n",
      "China Chef Restaurant ORG\n",
      "Pleasant Hill FAC\n",
      "6:45 pm TIME\n",
      "next Tuesday DATE\n",
      "6:30 pm TIME\n",
      "next Tuesday DATE\n",
      "2 CARDINAL\n",
      "China Chef Restaurant ORG\n",
      "a great day DATE\n"
     ]
    }
   ],
   "source": [
    "with open(\"/Users/tdubon/GitHub/Mastering-spaCy/Chapter10/data/Chapter10/data/utterances.txt\", \"r\") as sentences:\n",
    "    for sent in sentences:\n",
    "        doc = nlp(sent)\n",
    "        [print(ent.text, ent.label_) for ent in doc.ents]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "397e665c",
   "metadata": {},
   "source": [
    "## Intent Recognition"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5715a2de",
   "metadata": {},
   "outputs": [],
   "source": [
    "#see python notebook: \n",
    "#https://github.com/PacktPublishing/Mastering-spaCy/blob/main/Chapter10/Intent-classifier-char-LSTM.ipynb"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3aea27bb",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "050246df",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}