COLLOQUIUM Computer Science Department, Boston University Speaker: Prof. Ashish Goel Stanford University www.stanford.edu/~ashishg Title: Towards robust reputation systems Date: January 21, 2005 Time: 11:15 am Place: MCS 135 (for directions, see www.cs.bu.edu/colloquium) Abstract: We will study the robustness of reputation systems in several settings. Eigenvector based methods in general, and Google's PageRank algorithm for rating web pages in particular, have become an important component of information retrieval on the Web. These methods measure the reputation of a web-page as its stationary distribution in an associated random walk. Such methods are susceptible to collusions by groups of web-page owners interested in increasing the ranks of their web-pages. We will present metrics, hardness results, and algorithms for the problem of detecting colluding web-pages. In particular, we will define the amplification ratio of a set of web-pages as the amount of "extra reputation" this set generates for itself. We prove that the problem of finding sets with high amplification is formally hard, and then under certain assumptions on the mixing time of the web graph, obtain a polynomial time algorithm for identifying a colluding web-page. We then discuss the phenomena of ballot stuffing and bad mouthing in eBay like reputation systems. We prove that these phenomena can be mitigated provided the reputation premium is small compared to the transaction costs. This represents joint work with Hui Zhang, Kahn Mason, Ramesh Govindan, Benjamin Van Roy, and Rajat Bhattacharjee. Brief biography: Ashish Goel is an Assistant Professor of Management Science and Engineering and (by courtesy) Computer Science at Stanford University. He received his PhD in Computer Science from Stanford in 1999, and was an Assistant Professor of Computer Science at the University of Southern California from 1999 to 2002. His research interests lie in the design, analysis, and applications of algorithms. Professor Goel is a recipient of an Alfred P. Sloan faculty fellowship (2004-06), a Terman faculty fellowship from Stanford, and an NSF Career Award (2002-07). Host: John Byers