Trust Inference in Complex Trust-Oriented Social Networks

Abstract

Many social networking platforms have emerged on theWeb, such as MySpace and Facebook. In those networks,most participants are usually physically unknown in reallife and have no prior direct interactions with each other.Hence, it is necessary to form complex trust-oriented socialnetworks with more trust related information and infertrust values between participants. The inferred trust resultsprovide important indications for some applications, suchas introducing products to trustworthy people or recruitingtrustworthy employees from social networks. Traditionaltrust inference mechanisms are based on simple trust networksand can hardly deliver realistic trust results. In thispaper, we first propose a complex trust-oriented social networkstructure containing trust values, social relations andrecommendation roles. We also propose a novel Bayesiannetwork based trust inference mechanism taking trust values,social relations and recommendation roles into account.Experiments demonstrate that our proposed trust inferencemechanism performs well in complex trust-oriented socialnetworks.