This data download is a pre-trained model for a Bayesian classifier. If you do not have experience with Python NLTK, you may not be interested in this data set.

A 93.57% accurate Naive Bayes subjectivity classifier trained on IMDb plots and RottenTomatoes quotes. This classifier can be used for hierarchical sentiment analysis to determine whether text is objective or subjective before using a sentiment classifier to determine polarity. Given a feature dict of words and bigrams, it will classify the text as “quote” or “plot”, where “quote” means the text is subjective, and “plot” means the text is objective. The classifier requires Python & NLTK 2.0 and is licensed for commercial usage.

Application Gallery