Skip to content

Latest commit

 

History

History
4 lines (2 loc) · 416 Bytes

File metadata and controls

4 lines (2 loc) · 416 Bytes

Sentiment Analysis Using Bi-directional LSTMs

I make use of bi-directional LSTMs to perform sentiment analysis on movie reviews. The data was sourced through an IMDB movie review dataset [can be found in Kaggle]. The text data was preprocessed and then converted to vectors using one hot encoding. The resulting bidirectional LSTM had an accuracy of 87.37% in accurately identifying the sentiment of the review.