Different Feature Selection for Sentiment Classification

Sentiment Analysis (SA) research has increased tremendously in recent times .Sentiment analysis means to extract opinion of users from review documents. Sentiment classification using Machine learning (ML ) methods faces the problem of high dimensionality of feature vector. Therefore, a feature selection method is required to eliminate the irrelevant and noisy features from the feature vector for efficient working of ML algorithms Rough set theory provides an important concept for feature reduction called reduct. The cost of reduct set computation is highly influenced by the attribute size of the dataset where the problem of finding reducts has been proven as NP- hard problems.Different feature selection are applied on different data set, Experimental results show that mRMR is better compared to IG for sentiment classification, Hybrid feature selection method based on the RST and Information Gain (IG) is better compared to the previous methods. Proposed methods are evaluated on four standard datasets viz. Movie review, Product (book, DVD, and electronics) reviewed datasets, and Experimental results show that hybrid feature selection method outperforms than feature selection methods for sentimental classification.