Natural language processing
It is based on my learning of NLP which I did from online sources.
short_reviews is used to test and train the Classifiers.
Instead of using only one classifier, here I have used several classifiers like:
- Naive Bayes
- MNB
- BernoulliNB
- Logistic Regression
- LinearSVC
- SGDC
A vote from taken each of the algorithms and the sentiment which scored maximum votes was considered to be final.
Along with that a confidence value was also taken which was the ratio of the the positive votes or negative votes to the total number of votes.
sentiment_analysis.py is used to score and classify the sentiment as positive or negaive.