Analyzing and Comparing opinions on the Web mining Consumer Reviews

Product reviews posted at online shopping sites plays a major role in improving performance of various enterprises. To assess the performance, the posted reviews must be of good quality. The good quality is judged by using certain criteria (rules) to be satisfied. The criteria (rules) should be applied on the online reviews or the documents collected based upon reviews. Thus, it is considered to be very difficult for decision-maker with an efficient post processing step in order to reduce the number of rules. This project proposes a new classification based interactive approach to prune and filter discovered rules to eliminate low-quality reviews. The proposed approach to enhance opinion summarization is done in a two-stage framework which is (1) discriminates low quality reviews from high-quality ones and (2) enhances the task of opinion summarization by detecting and filtering low quality reviews. For the sentiment factor, we propose Sentiment PLSA (S-PLSA), in which a review is considered as a document generated by a number of hidden sentiment factors, in order to capture the complex nature of sentiments. Training an S-PLSA model enables us to obtain a succinct summary of the sentiment information embedded in the reviews.