Don't follow me: Spam detection in Twitter

摘要: The rapidly growing social network Twitter has been infiltrated by large amount of spam. In this paper, a spam detection prototype system is proposed to identify suspicious users on Twitter. A directed graph model explore the “follower” and “friend” relationships among Based Twitter's policy, novel content-based features graph-based are also facilitate detection. Web crawler developed relying API methods provided Around 25K users, 500K tweets, 49M follower/friend in total collected from public available data Bayesian classification algorithm applied distinguish behaviors normal ones. I analyze set evaluate performance system. Classic evaluation metrics used compare various traditional methods. Experiment results show that classifier best overall term F-measure. trained entire set. result shows can achieve 89% precision.