Trust-Based Recommender System in P2P Network

Distributed resource sharing in P2P networks has received a great deal of attention. Resources overload make users increasingly difficult to locate the right resource. Recommender systems have been proven to resolve this problem. But Recommender systems have natural weaknesses such as their vulnerability to various attacks. So, in this paper, we present a model of recommender system based on trust to figure this problem out. Firstly, we propose a computation model to compute trust value for each peer owning resources shared based on interactive histories or recommendations from other peers interacted. Then, right resources in right peers would be found for users based on trust value. So it can prevent some attacks from shilling of malicious peers. The experiment result has shown the reliability and robustness.