Thwarting P2P Pollution Using Object Reputation

This paper describes a distributed object reputation management scheme that counteracts content pollution in peer-to-peer filesharing systems. The proposed scheme allows honest peers to assess the authenticity of online content by securely tabulating and managing endorsements from other peers. We employ a novel voter correlation scheme to weight the opinions of peers, which gives rise to favorable incentives and system dynamics. We present simulation results indicating that our system is scalable, efficient, and robust to attack