Logic Programming for Multiagent Planning with Negotiation

Tran Cao Son, Enrico Pontelli, and Chiaki Sakama

Proceedings of the 25th International Conference on Logic Programming (ICLP'09), Lecture Notes in Computer Science 5649, Springer-Verlag, pages 99-114, 2009.


Multiagent planning deals with the problem of generating plans for multiple agents. It requires formalizing ways for the agents to interact and cooperate, in order to achieve their goals. One way for the agents to interact is through negotiations. Integration of negotiation in multiagent planning has not been extensively investigated and a systematic way for this task has yet to be found. In this paper, we develop a generic model for negotiation in dynamic environments and apply it to generate joint-plans with negotiation for multiple agents. We identify the minimal requirements for such a model and propose a general scheme for one-to-one negotiations. This abstract model of negotiation is then instantiated to deal with dynamic knowledge of planning agents. We demonstrate how logic programming can be employed as a uniform platform to support both aspects of planning and negotiating, thus providing an ideal testbed for experimenting with multiagent planning with negotiations.

Full Paper (520K) Slide (pdf 2965K) © Springer-Verlag (The original publication is available at www.springerlink.com)