Mining reputation of person/product from comment and reply on UCC/internet article

Abstract

Increasing Internet users and generalization of internet brought many changes to our Society. One of the most important changes is internet media start to provide their contents to people who have internet connection and these people express their opinions about UCC/Internet Articles as reply/comment after they read it. Replies/Comments written by users are from many different age groups which various opinions exist and these opinions are very useful for marketing in business or politician. However, collecting and analyzing opinions of users manually costs very high amount of money and a lot of labor. Therefore, this paper proposes a new method to extract opinions from reply or comment and to summarize extracted opinions to provide reputation of person/product in which we are interested. © 2010 Springer-Verlag Berlin Heidelberg.