Product Feature-based Ratings forOpinionSummarization of E-Commerce Feedback Comments

Abstract

the growth of internet, online social networking sites, blogs, discussion forums, etc have gained a tremendous importance. Consumers comment on net to express their views, feedbacks and opinions. The opinion of users is of great importance for mining useful information from the text which can be done through opinion mining techniques. Opinion mining or sentiment analysis is the computational field of study of people's opinions, emotions, and attitude towards particular Feature. When buying a new product buyer mostly refer the opinion of the other users who have bought the product. Hence, in this work a product Feature rating framework is being proposed. This dissertation comprises mainly of four modules preprocessing, Feature identification, review classification and Feature rating. Finally, the rating are been shown in the graph. For the analysis of the system, we have used Amazon review dataset which consists of customers reviews about product. In the system Apriori algorithm is used for Feature identification, Support Vector Machine algorithm for review classification and SentiWordNet lexicon for giving rating to each Feature of the product. KeywordsMining, Sentiment Analysis, Feature