ryrutherford / sentiment-analysis

A supervised ML project to identify the best performing model on sentiment analysis in movie reviews. I used a set of 5331 positive sentences and 5331 negative sentences from IMDB reviews to train and test the models.

Sentiment Analysis of IMDB Movie Reviews (Supervised Learning Project)

• I use scikit-learn, nltk, and NumPy to train and test different ML classifiers on 10,000+ sentences reflecting either positive or negative sentiment of a movie
• The goal of this project is to determine which model and what pre-processing (stemming, lemmatization, stop words, minimum frequency) are most effective at predicting positive or negative sentiment from a single sentence