Sentiment analysis on movie reviews by recurrent neural networks and long short-term memory

Abstract

Sentiment analysis has become important tool that can analyse review on any product or service that can be reviewed. Same goes to movie, all the audient are freely to make their own reviews on the movie that they watch and the reviews can be positive or negative based on audient satisfactions. Automated sentiment analysis is very important to make sure the analysis produce an accurate result and in faster time. By using the deep learning as the based to create the automated sentiment analysis it will be the great decision because of the deep learning structure that have multilevel of layer that can have sensitive process to classify the data. Upgrading the sentiment analysis using Recurrent Neural Networks (RNNs) and addition of Long Short-term Memory (LSTM) and also some modification on the number of layer with the mathematical calculation can improve the analysis accuracy. The dataset of the movie reviews will be collected on IMDB movie reviews database.