期刊名称：International Journal of Innovative Research in Science, Engineering and Technology
印刷版ISSN：2347-6710
电子版ISSN：2319-8753
出版年度：2015
卷号：4
期号：12
页码：12196
DOI：10.15680/IJIRSET.2015.0412091
出版社：S&S Publications
摘要：Consumers used to review regarding many products which are available on-line. These reviews arevaluable and important for both firms and users. But many times it seems that these reviews are often disorganizedmaking it difficult to do information navigation and knowledge extraction. So here product aspect ranking frameworkis proposed so as to identify the important aspects and improve the usability of reviews. There are 2 ways for importantproduct aspect identification as: 1) they are commented on by a large number of consumers and 2) customer's opinionson the important aspects influence their overall opinions on that particular product. In particular, if reviews of theproduct are given by the customer, initially it identifies the product aspects by a standford parser then determinesconsumer opinions via sentiment classification. After that probabilistic aspect ranking algorithm will be used.
