What and why?

Being able to extract the sentiment polarity (positive, negative, or neutral) of free text can have many important applications. Here I am comparing three open-source sentiment classifiers for Python and performing a quick quality check.


Set #1

Set #2

Set #3


For set #1 ("This is the best movie!"), all three classifiers got similar results as all pointed towards positive sentiment as expected.

Interestingly enough, for set #2 ("This is a great movie!"), Senti_sentiment classifier seemed to have difficulty classifying that as a positive text.

Lastly, for set #3 ("I don't feel good"), Vader sentiment classifier was the only one that correctly signaled a negative sentiment, by taking the negation of a text into consideration.

Please rate this