Bayesian Learning in Negotiation y

Recent growing interest in autonomous interacting software agents and their potential application in areas such as electronic commerce (Sandolm & Lesser 1995) has given increased importance to automated negotiation. Much DAI and game theoretic research (Rosenschein & Zlotkin 1994; Osborne & Rubinstein 1994) deals with coordination and negotiation issues by giving pre-computed solutions to speci c problems. There has been much research reported on developing theoretical models in which learning plays an eminent role, especially in the area of adaptive dynamics of games (e.g., (Jordan 1992; Kalai & Lehrer 1993)). However, to build autonomous agents that improve their negotiation competence based on learning from their interactions with other agents is still an emerging area. We are interested in developing autonomous agents capable of reasoning based on experience and improving their negotiation behavior incrementally. Learning in negotiation is closely coupled with the issue of how to model the overall negotiation process, i.e., what negotiation protocols are adopted. Standard gametheoretic models (Osborne & Rubinstein 1994) tend to focus on outcomes of negotiation in contrast to the negotiation process itself. DAI research (Rosenschein & Zlotkin 1994) emphasizes special protocols articulating compromises while trying to minimize the potential interactions or communications of the involved agents. Since we are motivated by a di erent set of research issues, such as including e ective learning mechanisms in the negotiation process, we adopt a di erent modeling framework, i.e., a sequential decision making paradigm (Bertsekas 1995; Cyert & DeGroot 1987). The basic characteristics of a sequential decision