P2P networks have become increasingly popular in the recent years. However, their open, distributed and anonymous nature makes them very vulnerable against malicious users who provide bad responses to requests from other peers. Motivated by this observation, various solutions for distributed reputation systems have been presented recently. In this paper, we describe the first reputation system which incorporates both user-individual personalization and global experiences of peers in the network, for the distributed computation of reputation values. We also present a secure method to compute global trust values, thus assuring identification and isolation of malicious peers. Finally, our simulations show that our system is robust even against attacks from groups of malicious peers deliberately cooperating to subvert it.
Paul-Alexandru Chirita, Wolfgang Nejdl, Mario T. S