This project demonstrates sentiment analysis in Python using two different techniques:
- VADER (Valence Aware Dictionary and sEntiment Reasoner) - Bag of words approach
- Roberta Pretrained Model from Huggingface Pipeline
- Reading data using pandas
- Basic NLTK operations like tokenization and part-of-speech tagging
- Using NLTK's SentimentIntensityAnalyzer to get sentiment scores
- Applying VADER on the entire dataset
- Visualizing VADER results
- Utilizing a pretrained model from Huggingface Transformers
- Comparing scores between VADER and Roberta model
- Visualizing scores between models
- Pairplot to compare VADER and Roberta scores
- Exploring examples where model scoring and review score differ
- Showing examples of positive 1-star and negative 5-star reviews
- Demonstrating the Transformers Pipeline for quick sentiment predictions
- Ensure all dependencies are installed (
pandas
,numpy
,matplotlib
,seaborn
,nltk
,transformers
) - Download or clone the notebook and associated dataset
- Run each cell in the notebook sequentially
- pandas
- numpy
- matplotlib
- seaborn
- nltk
- transformers
Make sure to replace tutorial_link_here
in the README with the actual link to the tutorial.