SENTIMENT ANALYSIS OF MOVIES REVIEWS USING IMPROVISED RANDOM FOREST WITH FEATURE SELECTION

Abstract

Human psychology has perpetually influenced by means of others suggestion and experiences. Our experiences about whatever are very a lot influenced by means of different's experiences, and at any time when we have to make a choice or answer, we regularly search out different's reports, so folks are excited to understand other's experiences for his or her profit because of this computerized sentiment analysis systems are required. Sentimental evaluation, also known as opinion mining, is a ordinary language processing system used to extract the feeling or perspective of common lots regarding a given subject or product.This paper proposes a prediction model for the sentiment analysis of movies review using improvised random forest classification algorithm. The proposed work consists of hybrid approach of feature selection with classification for prediction analysis in text mining. Gini Index feature selection technique is used for optimizing the features. The experimental results of the proposed model are evaluated with respect to existing techniques and showed that the proposed technique achieved 90.69% accuracy and precision, recall class parameters of propose technique are also better than the existing techniques.