Archive | 2019
Improved Feature Based Sentiment Analysis for Online Customer Reviews
Abstract
The evolution of E-commerce site tends to produce a huge amount of data nowadays. These data consist of very novel knowledge to compete with the other E-commerce sites in terms of business perspective. Customers often use these E-commerce sites to manage decision on the purchase of products based on comments or reviews given by the existing customer who bought the same product. The concept of Opinion Mining enables these processes of selection and decision easier. Several techniques have been proposed for the opinion mining and provided their own advantages. However, those techniques contain certain drawbacks in the selection of features and opinions with respect to the priority of product feature given by the individual user. This paper proposes a novel idea of incorporating weight which is automatically calculated according to the attributes evolved. The reason to do certain weight calculation is an assumption of weight and weight based on priority may differ from person to person. Experimental results show the performance of the proposed with various techniques for the online review collected from different sites.