A sequential model of bargaining in logic programming

This paper proposes a sequential model of bargaining specifying reasoning processes of an agent behind bargaining procedures. We encode agents’ background knowledge, demands, and bargaining constraints in logic programs and represent bargaining outcomes in answer sets. We assume that in each bargaining situation, each agent has a set of goals to achieve, which are normally unachievable without an agreement among all the agents who are involved in the bargaining. Through an alternating-offers procedure, an agreement among bargaining agents may be reached by abductive reasoning.We show that the procedure converges to a Nash equilibrium if each agent makes rational offers/counter-offers in each round. In addition, the sequential model also has a number of desirable properties, such as mutual commitments, individual rationality, satisfactoriness, and honesty.