The optimization model of trust for white-washing

Abstract

With the increase of trust recommendation in the network services and applications, a large number of corresponding attacks including whitewashing attacks have emerged. This paper proposes a trust optimization model in recommendation system to counter the white-washing attacks. The proposed model first separates the white-washing nodes from other nodes by dividing the nodes into different groups. The grouping will limit the dubious white-washing nodes in the proposed model so as to take these nodes much more cost to become a normal node. Meanwhile, the normal interactions are not affected. Our experimental results show that the proposed model can insulate suspicious white-washing nodes effectively, and ensures the normal interaction activities among normal nodes at the same time. Thus, the proposed model can resist the white-washing attacks effectively.