Enabling dynamic roles for agents

Abstract

Sociality, coordination and interaction play a fundamental role in today's complex systems and applications. Role-Based Collaboration (RBC) is one of the best methodologies to facilitate and manage complex organizational structures through interactions and sociality artifacts. RBC is adopted in a lot of scenarios including Multi Agent Systems (MASs), where agents are playing effectively as human being counterparts. Today's complex systems require a great degree of dynamism and flexibility in order to improve sociality and interaction management of involved entities, and therefore every RBC implementation must face such needs. In this paper we present the WhiteCat framework, a role based agent system that provides an high degree of flexibility and dynamism in role assumption, use and release. Moreover, WhiteCat introduces a few innovative features over the RBC methodology taking inspiration by real life, as for instance the role perceivability, that is the capability of an entity to recognize the role played by a counterpart without having to query a yellow page system or something similar, but simply analyzing the counterpart itself.