Preferring and Updating in Abductive Multi-agent Systems

We present a logical framework and the declarative semantics of a multi-agent system in which each agent can communicate with and update other agents, can react to the environment, is able to prefer, whether beliefs or reactions, when several alternatives are possible, and is able to abduce hypotheses to explain observations. The knowledge state of an agent is represented by an updatable prioritized abductive logic program, in which priorities among rules can be expressed to allow the agent to prefer. We sketch two examples to illustrate how our approach functions, including how to prefer abducibles to tackle the problem of multiple hypotheses and how to perform the interplay between planning and acting. We argue that the theory of the type of agents considered is a rich evolvable basis, and suitable for engineering configurable, dynamic, self-organizing and self-evolving agent societies.