An actor network-based approach to pirates community discovery in peer-to-peer network

Abstract

Due to the potential content sharing applications in civilian environment, peer-to-peer (P2P) network has received considerable attention. Unfortunately, the problem of pirates widely exists in P2P network as the absence of content supervision. Piracy propagation shares high similar piracy contents frequently to constitute the pirates community. Therefore in the P2P network, pirate community discovery is one of the most important research topics. Interest-based and random walk-based approaches have been developed for community discovery in the P2P network, but they work ineffectively in pirates community discovery. Because they fail to simultaneously consider the message similarity and the message passing frequency. This paper argued that these factors should not be ignored in the pirates community discovery. Based on the analysis, we proposed an actor network-based approach and extended the Girvan-Newman (GN) algorithm for pirates community analysis. The experimental evaluation confirms the effectiveness of the approach. © Springer International Publishing Switzerland 2013.