A Sentiment module made with multiple classifiers for accuracy and vote of confidence being voted among the classifiers in order to get high accuracy.
Accuracy : 72.74%
The dataset of the sentiments has been attached in the repository along with pickles of the classifiers used.
SklearnClassifier, MultinomialNB, BernoulliNB, LogisticRegression, SGDClassifier, SVC, LinearSVC, NuSVC.
NLTK, Pickle, Sklearn, Random, Statistics.