Managing Reputation in Collaborative Social Computing Applications

Abstract

Reputation is a fundamental concept for making decisions about service providers. However, managing reputation in peer-to-peer distributed applications is not easy due to the lack of a central server that can compute this property from user opinions. Moreover, users have to marry this information with their individual trust in the service provider, which may be based on their past experiences, the opinions of their direct contacts, or both. This paper develops a reputation management system embedded in the Digital Avatars framework for collaborative social computing applications, using subjective logic. We show how the reputation of a given service provider can be calculated using the users’ opinions about it, and how this reputation can be explicitly represented, managed and combined with the trust that individual service requesters may have in them, in order to make better informed decisions