Anticipating Agents' Negotiation Strategies in an E-marketplace Using Belief Models

In this paper we present a Markov Decision Process approach which gives agents the ability to learn about other agents' strategies from past negotiation steps. Over time, the agents can update their belief models regarding other agents and use those models to anticipate their own strategies during a negotiation process. The negotiation model described in this paper is devised for an agentbased marketplace for personnel acquisition. We briefly describe a number of tactics which agents can employ during negotiations. An agent may change these tactics over time to avoid the shortcomings of a deterministic and fixed negotiation strategy by adaptively deciding which tactic will be the best one for the next negotiation step.