Linkage liaisons: a scenario for the computational study of conflict and structure in multiagent systems

Abstract

Website presentation is becoming a crucial issue as more and more services of all different kinds are offered by more and more content providers to a rapidly growing audience on the Internet. In the face of accelerated competition and intransparency it is highly important for a service provider to present himself with an appealing and professionally designed website. To attract the attention of as many Internet surfers as possible and to raise their curiosity and interest, service providers will have to do everything to enhance their website’s traffic and to transform casual visitors into habitual customers and partners. In order to attract more traffic, linkages directing the user from website to website is a crucial issue and building-up strategic linkage liaisons with other website owners on a B2B basis will be one of the future key elements of successful website presentation. Linkage liaisons between websites will give rise to emergent network structures (professional and non-professional) as well as to all sorts of minor and major linkage conflicts that arise among website owners (or their agents) in the course of incrementally engaging in building up new linkages and deleting old ones. Hence, linkage liaisons look like a particularly suitable scenario for designing, implementing, and testing models for real world applications and for scientific research purposes of comparing interactionism with social systems approaches in sociology. We assume that liaison models differ from more conventional DAI applications not only because they operate on a permanent basis (cf. computational ecology) and that they produce, reproduce and modify an ”emergent” social structure while resolving (or not resolving!) conflictive episodes. The presumed advantages of the proposed model, both from a software engineering and from a sociological perspective, should be seen in the way that a linkage networks would have to meet the demands of shifting from a highly efficient and cost saving routine mode into a resource consumptive conflict mode and back again into a new routine mode after the conflict has been settled. In other words: the system should be able to learn from conflicts and, in doing so, an agent society would operate like a self-organising mechanism or a selfsustaining general purpose infrastructure.