The Design of Utility Functions for Information E-marketplaces with Price Quo te Fees
Itai Yarom, Claudia V. Goldman, Jeffrey S. Rosenschein

Abstract

Information e-marketplaces enable entities to buy and sell information; these buying and selling entities can be humans, or automated agents that represent them. In this paper, we analyze the characteristics of information e-markets and the behavioral character of the autonomous agents that operate in these markets. We describe four desirable characteristics of information sellers' utility, and present a specific definition of utility that exhibits these characteristics.

We continue by addressing the bandwidth problem that sellers face due to buyers and sellers that perform price sniffing. We show that this argues for a system that includes middle-agents along with buyers and sellers. Furthermore, sellers and middle-agents can protect themselves from the costs associated with multitudes of requests for price quotes (RPQs) by establishing a fee for those requests. This forces buyers and sellers to consider more carefully what prices are truly important to know, which in turn leads to the introduction of intelligent agents that will decide when it is profitable to look for additional price quotes.