MINING FEATURE-OPINION FROM CUSTOMER REVIEWS FOR SENTIMENT CLASSIFICATION

Abstract

Almost all people want to gain more and more information about the products, before they purchase them. Therefore, they ask their friends, search on net and then decide to buy a product. As there is tremendous increase in e-commerce, almost every company provides a customer feedback data form on its website. Many sites put stress on participation of users, more and more Websites, such as Amazon, UCI lead people to write their opinion about products they are interested in. So, the number of product reviews from customer is also increasing. The opinion not only helps the customer to buy good product, also help the product manufacture to see the pros and cons of their product and also show the comparison of his product with the other competitor and help the product manufacturer to see which product is liked/disliked by the customer and they can improve their product future. Therefore it becomes very difficult for manufacturers to analyze every review for analyzing the product. Hence, we have made a system that takes customer reviews and finds positive and negative feature with their semantic orientation. We have used POS tagging for each sentence, and then extracted the features from the customer reviews. We have then worked on co-reference (pronoun) resolution before summarizing the semantic orientation for all features. First we have find features of the product and then reduced them using Word Net Similarity for grouping similar features. Finally, we will classify as positive, negative or neutral, so that a customer gets a better idea of each particular feature of a product and can compare with other products.