Agents advanced features for negotiation in Electronic Commerce and Virtual Organisations formation process

Abstract

Electronic commerce technology has changed the way traditional business is being done. Transaction complexity is increasing due both to the huge amount of available information and also to the environment dynamics. Moreover, electronic commerce has enabled the arising of new economical structures, as in the case of virtual organisations. Our research aims at providing flexible and general-purpose systems for intelligent negotiation, both for electronic commerce and virtual organisation formation. This paper proposes an electronic market architecture implemented through a multi-agent system. This architecture includes both a specific market agent which plays the role of market coordinator, as well as agents representing the individual business partners with their own goals and strategies. We also include a sophisticated negotiation protocol through multi-criteria and distributed constraint formalisms. An online, continuous reinforcement learning algorithm has been designed to enable agents to adapt themselves according to the changing environment, including competitor agents.