Sentiment Analysis of IMDb Movie Reviews Using Long Short-Term Memory
Saeed Mian Qaisar
Abstract
The sentiment analysis is an emerging research area where vast amount of data are being analyzed, to generate useful insights in regards to a specific topic. It is an effective tool which can serve governments, corporations and even consumers. Text emotion recognizing lays a key role in this framework. Researchers in the fields of natural language processing (NLP) and machine learning (ML) have explored a variety of methods to implement the process with highest accuracy possible. In this paper the Long Short-Term Memory (LSTM) classifier is used for analyzing sentiments of the IMDb movie reviews. It is based on the Recurrent Neural Network (RNN) algorithm. The data is effectively preprocessed and partitioned to enhance the post classification performance. The classification performance is studied in terms of accuracy. Results show a best classification accuracy of 89.9%. It confirms the potential of integrating the designed solution in modern text based sentiments analyzers.