Feature level sentiment analysis on movie reviews

Abstract

Sentiment Analysis is a very challenging task in these days, which is a great requirement in every field as in Political field, Marketing field and in social field mainly. Many researchers have been taken a lot of interest in this area and proposed their work. The Sentiment analysis is done on the basis of Document, Sentence and feature levels. But the first two levels didn't consider object features that have been commented in a sentence. So the feature level sentiment analysis is more appropriate compare to both. Different types of tools and approaches have been used by the researchers for pre-processing, tagging, semantic orientation, and finally for calculating scores for deriving sentiments of the reviews. But the problems, found in analysis generally are, where reviews contain the negative, intensifier, conjunctive and synonyms words. And other problems are coreference resolution; anaphora resolution, named-entity recognition, and word-sense disambiguation. These problem are unresolved sometimes so that the performance of sentiment analysis decrease. A lot of work is done mainly in Product, News, Sports domains. We proposed our work in Movie Reviews because it's a more attractive area, for now-a-days generation where multiple sites allow users to submit reviews describing what they either liked or disliked about a particular movie. In this work we have proposed a system which classifies the polarity of the movie reviews on the basis of features by handling negation, intensifier, conjunction and synonyms with appropriate pre-processing steps. We have used SentiWordNet tool for calculating the scores of reviews.