Overview

Reputation systems are becoming an increasingly important component of information retrieval on the Web. Such systems are now ubiquitous in electronic commerce, and enable users to judge the reputation and trustworthiness of on-line merchants or auctioneers. In the near future, they may help counter the free-rider phenomenon in peer-to-peer networks by rating users of these networks and thereby inducing social pressure to offer their resources for file-sharing. Also, they may soon provide context for political opinion in the Web logging (blogging) world, enabling readers to calibrate the reliability of news and opinion sources.

In keeping with the design philosophy of many of these system, we consider the design of distributed rating systems. As a case study, we illustrate two different approaches to a distributed rating system aimed at tackling the free-rider problem in P2P networks. A key challenge in designing such rating schemes is to make them collusion-proof. We study this problem in the context of eigenvector-based reputation systems.

People

• Hui Zhang
• Debojyoti Dutta
• Ashish Goel (Stanford)
• Ramesh Govindan
• Kahn Mason (Stanford)
• Benjamin Van Roy (Stanford)

Publications

• Debojyoti Dutta, Ashish Goel, Ramesh Govindan, Hui Zhang.The Design of A Distributed Rating Scheme for Peer-to-peer Systems. the first Workshop on Economic Issues in Peer-to-Peer Systems, Berkeley, CA (June 5-6, 2003).
• Hui Zhang, Ashish Goel, Ramesh Govindan, Kahn Mason, and Benjamin Van Roy.Making Eigenvector-based Reputation Systems Robust to Collusion. the Third Workshop on Algorithms and Models for the Web Graph, October 2004.

Talks

• Hui Zhang. Making PageRank Algorithm Robust to Collusion. at Google Inc, April 2004.