Modeling agents' knowledge in collective evolutionary systems

Abstract

Collective evolutionary systems have drawn considerable attention in recent times. One reason for this is the advances in the second generation of web development and design. This has led to the emergence of web-based communities and applications such as social-networking sites and blogs. The collective behavior of the system depends heavily on the performance and the reasoning aspects of the agents. Although logics of knowledge has been extensively studied, its application to domains in collective systems has not been illustrated before. In this paper we analyze some significant evolutionary domains characterized by collective and evolutionary aspects, such as partial observability, distribution, sharing, coordination, and mobility. We show how knowledge in these domains can effectively be modeled using ELMA (epistemic logic for mobile agents), which is an extension of an existing epistemic logic with the notion of space and containment, that entails the concepts of group and collaboration. The need and features of appropriate reasoning and planning mechanisms for collective evolutionary environment are also discussed. © 2009 Springer Berlin Heidelberg.