Naive Bayes Classifier

[honnow]

hi,when I used the Naive Bayes Classifier function in KH Coder software,i want to do sentiment analysis,but it happens all the time.

i don't know how to solve it ,please give me some guidance.thanks.

[ko-ichi-h]

Please fill in the template below:

About this particular case, the most important thing is that you attach the input files.

Version number of KH Coder
To check the version number, go to "Help" "About" in the menu of KH Coder. Please be sure to enter it correctly, such as "3.Beta.07f", not just "3".

OS
e.g. Windows 10, Ubuntu 20.04

The error message as detailed as possible
Please do not abbreviate or translate the message. Please tell me as it is, verbatim. If you don't like to type, you can attach a screenshot.

To Reproduce
Steps to reproduce the behavior:

1. Go to '...'
2. Click on '....'
3. Scroll down to '....'

Expected behavior
A clear and concise description of what you expected to happen.

What happens if you perform the same operation with the tutorial file (Anne.xls)
Can you get the intended result with tutorial data files? Or you get the same error? Please use "Anne.xls" in "tutorial_en" folder to test it out.

Files to Reproduce the Error
If we can't reproduce the error with the tutorial file "Anne.xls", please consider attaching files that can reproduce the Error.

Language of the Text
What language of text are you trying to analyze?
e.g. English, Spanish

Type of the data file
What kind of file are you trying to analyze?
e.g. Plain text (.txt), Excel (.xlsx), Word (.docx)

Screenshots
If applicable, add screenshots to help explain your problem.

CPU
e.g. Core i7, Atom

Antivirus or Security software running on the PC
e.g. Norton 360, Avast

Additional context
Add any other context about the problem.

Contents of the console window
Other than errors in pop-up windows, the full console/terminal output really helps. On Windows, you can find the console window by clicking KH Coder icon in the task bar. The window has white texts on a black background.

If it is long, you can save the contents in a text file, Word file, etc. and paste the file. Alternatively, you can attach a screenshot of the console window, but if possible, a copy and paste of the entire content is more likely to give us a clue.

[honnow]

May I ask, when conducting sentiment analysis on comments from different stages of an event, is it necessary to model each stage’s comments separately?

[honnow]

kappa and the correct classification ratio is diffrent,The two values are far apart。how to explain.thanks.

[ko-ichi-h]

Have you Googled “Naive Bayes Kappa”?

https://www.hindawi.com/journals/mpe/2014/383671/

Kappa measures the agreement between two raters who each classify items into mutually exclusive categories. Kappa is computed as formula (19), is the observed agreement among the raters, and is the expected agreement; that is, represents the probability that the raters agree by chance. The values of Kappa are constrained to the interval . Kappa = 1 means perfect agreement, Kappa = 0 means that agreement is equal to chance, and Kappa = −1 means “perfect” disagreement:

For data with such a distribution, if you answer "all -1", the correct answer rate will be high. Even if the correct answer rate is high, if it is close to the chance range, the Kappa value will not be high.