We decided to read about reputation for next week. Here are some papers I'm suggesting, both dealing with reputation.
Resnick, Paul and Richard Zeckhauser (2002). Trust Among Strangers in Internet Transactions: Empirical Analysis of eBay's Reputation System. The Economics of the Internet and E-Commerce. Michael R. Baye, editor. Volume 11 of Advances in Applied Microeconomics. Amsterdam, Elsevier Science.
Toshio Yamagishi Improving the Lemons Market with a Reputation System: An Experimental Study of Internet Auctioning
Kollock, Peter. The production of trust in online markets. In Advances in Group Processes (Vol. 16), edited by E. J. Lawler, M. Macy, S. Thyne, and H. A. Walker. Greenwich, CT: JAI Press. 1999.
Hess and Hagen, 2002, Informational Warfare
Some questions to think about: How can we compare online reputation systems, in which the information about the subject is attached to the subject, with gossip based reputation systems, in which the information about the subject passes through a social network, but is not part of the subject? Can we design a network/gossip based system for online communities that would overcome some of the ebay style system drawbacks? How can we model the difference between the implicit and explicit social sanctioning?
If you have a paper you'd like to recommend we read this week, please do so.
Here is a big selection of papers about reputation.
As far as sanctions go: it is worth considering a reputation system that allows only positive feedback, never negative, especially if the information is attached to the subject.
People in a group of sufficient size will always seek ways to abuse the system to harrass or punish others inordinately. Disallowing explicit sanctions prevents people from even attempting this behavior. Do sanctions provide enough social benefit that it is worth risking their abuse?
Posted by: Andrew Fiore at June 24, 2003 04:16 PMAlso: we've been talking about explicit reputation. We might want to look also at implicit reputation, that which derives solely from the behaviors of the person in question and others' reaction to them. In a newsgroup, reactions are manifested as replies (or lack of replies).
Can a bunch of characteristics and reactions be put together in a weighted combination that gives some idea of the person's "reputation"?
I wrote a paper about this for a class a couple of years ago if anyone is interested:
Creating and evaluating behavioral descriptors of newsgroup participants