SENTIMENT ANALYSIS USING BI-LSTM AND RECURRENT NEURAL NETWORK IN TWITTER DATASET

Abstract

The feeling of web user has a great influence on rest of the users, product sellers and market analysis. It is necessary to structure the unstructured data from various social platforms for proper and meaningful analyses. For the classification of   multi-language data, the analysis of feeling has recognized significant attention. This is called text organization that may be used to classify state of mind or feeling expressed in different ways like: positive, negative, favourable, un-favourable, thumbs up, thumbs down, etc. in the field of Automatic Language Processing. Because of machine learning ability, deep learning models are effectively used for this purpose. We propose solutions to sentimental analysis problem by implementing algorithms and to contract the result, we compare precision factor to find the best solution for sentimental analysis.