Graph-based detection of shilling attacks in recommender systems

Collaborative filtering has been widely used in recommender systems as a method to recommend items to users. However, by using knowledge of the recommendation algorithm, shilling attackers can generate fake profiles to increase or decrease the popularity of a targeted set of items. In this paper, we present a method to make recommender systems resistant to these attacks in the case that the attack profiles are highly correlated with each other. We formulate the problem as finding a maximum submatrix in the similarity matrix. We search for the maximum submatrix by transforming the problem into a graph and merging nodes by heuristic functions or finding the largest component. Experimental results show that the proposed approach can improve detection precision compared to state of art methods.