Publications‎ > ‎

Unbiased Trust Estimation in Content-Oriented Social Networks


Online communities based on user generated content (UGC) often rely on a small set of highly loyal and productive users to identify or create content that would interest their broader audience. Through continual contact, this user base often develops informal reputational and social ties among themselves. Websites often encourage the formalization of these trust networks by providing various social networking features, to increase engagement and loyalty. A user may come to trust a subset of the others as a consistent source of good content, while avoiding the remainder, leading to influence-based fragmentation. We discuss the impact that these emergent social behaviors can have on the quality of reputation scores, and develop algorithms that are able to take them into account while calculating user reputation.

S. Budalakoti, D. DeAngelis, and K. S. Barber. Unbiased Trust Estimation in Content-Oriented Social Networks. In the Proceedings of The Workshop on Trust in Agent Societies at The Tenth International Conference on Autonomous Agents and Multiagent Systems (AAMAS-2011); Taipei, Taiwan; May 2-6, 2011.

Dave DeAngelis,
Apr 29, 2011, 2:22 PM