E-business Websites Evaluation Based on Opinion Mining

With the popularization and development of Internet, as a new business methodology, e-business Web sites are attracting the attention of public and springing with fierce competition. How to make the Web sites invincible and how to help the potential customers make decisions? The only way is to make a comprehensive and accurate evaluation of the sites. Many traditional approaches have done this job, which are mostly qualitative and subjective. In this paper, we propose a new approach based on opinion mining. Through the reviews customers have posted we can mining the evaluation of various Web site indexes quantitatively. To improve the accuracy of the mining results we use a approach called MRA (mutual reinforcement approach). Different from the vast majority of opinion mining methods centering on mining explicit factor-opinion pairs, the MRA can mining implicit pairs effectively based on a pre-constructed association set. The biggest flash point of our evaluation approach is that it is customer-oriented and therefore more credible and trustworthy than traditional approach. The experimental study and evaluation show that the results by the proposed approach are consistent with the traditional approach but more reliable.