Accuracy of metrics for inferring trust and reputation in semantic web-based social networks

Abstract

While most research on the topic of trust on the semantic web has focused largely on digital signatures, certificates, and authentication, more social notions of trust which are reputation-based are starting to gain attention. In this paper, we describe an algorithm for generating locally-calculated reputation ratings from a Semantic Web Social Network. We present mathematical and experimental results that show the effectiveness of this algorithm to accurately infer the reputation of a node. We then describe TrustMail, an application that uses the network for rating email. © Springer-Verlag Berlin Heidelberg 2004.