Attack Detection by Rough Set Theory in Recommendation System
Famei He, Xuren Wang, Baoxu Liu · 2010
Recommender systems are widely used in e-commerce and information processing fields. But security problems arise at the same time. Recommender systems are vulnerable to profile injection attacks, by which malicious users add biased ratings into the rating database in order to change the recommedation results of certain items. This paper deals with classification and detection of the profile injection attacks with rough set theory. Empirical results illustrate that the method can detect profile injection attacks more accurately compared with prior works.