THE POTENTIAL FOR THE EVOLUTION OF COOPERATION AMONG WEB AGENTS

In building intelligent network agents, computer scientists may employ a variety of different design strategies, and their design decisions can have a significant effect on the ultimate nature of network intera ctions. Some agent designs are ''co-operative'', and populations of ag ents based on them would be able to interact smoothly, effectively uti lizing network resources. In contrast, other agent designs can lead to ineffective and wasteful competition for network resources, resulting in massive bottlenecks and unacceptable access delays. We focus here on a particular design question, the multiple-access problem: if an ag ent seeking a piece of information knows of several sites that have, o r might have, that information, how many queries should it issue, and when? We provide a formal analysis that demonstrates the viability of co-operative responses to this question under certain assumptions. We then discuss the limitations of this analysis and present the results of experiments done using a genetic-algorithms approach in which simul ated network agents ''evolve'' co-operative strategies, under less res trictive assumptions than those made in the formal analysis. (C) 1998 Academic Press Limited.