Movie review sentiment analysis is a complex problem under the scope of natural language processing. The huge variations in sentiments while expressing reviews make it very challenging. In this work, we have made an effort to deal with movie reviews having 5 different classes of sentiments namely "negative", "somewhat negative", "neutral", "somewhat positive", "positive" in increasing order of positivity. A sentiment analysis framework named MoReS is proposed which uses the power of deep learning to classify reviews on the basis of their sentiments. MoReS uses a combination of several components like data preprocessing, tokenization, embedding and a trained bidirectional Long Short Term Memory (LSTM) network to perform the classification. MoReS has been tested on a famous Rotten Tomatoes dataset hosted on Kaggle. From the experimental outcomes, it has been confirmed that MoReS is an efficient tool for sentiment analysis.
