Spammer Detection Based On Behavior And Relationship Of Users In Sina Weibo

The Internet has kept a rapid development since its birth and become an indispensable part of modern life. One of the most important problems and challenge the Internet service faced and grown with is the spam. Weibo (a typical microblogging service), an emerging social network service, also has a variety of spam problems. Aiming at this problem, this paper mainly focuses on the following three aspects:Firstly, the Internet spam problem, especially in Weibo, and the current Weibo spam detection research are described and summarized.Secondly, for the aggressive following to get followers behavior of spammer, a spammer growing model is presented and the increase of followers’number is predicted using probabilistic relational models (PRMs).At last, a method of historical consistency clustering is used to detection spammer in Sina Weibo which has82%accuracy rate and compared with method of user’s features and behavior using PRM. The spammer’s features also are analyzed.