A new approach towards co-extracting opinion-targets and opinion words from online reviews

Abstract

With the speedy expansion of e-commerce, more and more products are sold on the Web, and so many people are also purchasing products online. In order to enhance customer satisfaction and shopping experience, it has become a common practice for online merchants to enable their customers to review or to express opinions on the products that they have purchased. This work also displays an investigation of existing co-extracting algorithm and models are utilized to concentrate opinion targets and opinion words. Next, a graph-based co-ranking algorithm is used to extract opinion targets and opinion words. Also we are going to calculate relations between words, such as topical relations, in Opinion Relation Graph using TWTM (Topical Word Trigger Model). TWTM models topic specific word triggers, which are more discriminative. Hence TWTM is able to bridge the vocabulary gap between document content and key phrases more precisely.