Constraint Based Automated Multi-attribute Negotiations

Multi-attribute (Multi-issue) negotiation protocols have been studied widely, and represent an important challenge in the multiagent systems community (Lai et al., 2004). Therefore, a lot of automated negotiation models and protocols have been developed, and manifold negotiation challenges have been already addressed. Most research in automated negotiation to date has focused on the competitive aspect (Vo et al., 2007). On the other hand, work by Dispute Resolution theorists in the social sciences has also focussed substantially on how to achieve negotiated agreements that are of a high value to all parties (Fischer & Ury, 1981). This approach is known as Integrative or Interest-based negotiation, and it has been recognised as the more successful approach to the negotiation problem. Example scenarios where such cases may arise are: business process management involving agents within the same organization, e-commerce negotiations where the seller is interested in having a satisfied buyer (e.g. long-term commercial relationships), or e-commerce scenarios where risk averse agents avoid the conflict in the negotiation processes. In the context of purchase negotiation scenarios, it is clear that every negotiation partner tries to maximize his preferences. However, when an agent aims at optimizing his own benefit with no regard for the others’, it has been shown that negotiators more often than not reach inefficient compromises. Conflict theorists Lax and Sebenius (Lax & Sebenius, 1992) argue that negotiation necessarily includes both cooperative and competitive elements, and that these elements exist in tension. Therefore, he refers to the problem of deciding whether to pursue a cooperative or a competitive strategy at a particular time during a negotiation as the Negotiator's Dilemma. However, it is not always possible to separate the integrative bargaining process, i.e. when agents use cooperative strategies to search for joint gains, from the distributive bargaining process, i.e. when agents use competitive strategies in order to 'claim value'. The main problem is that distributive and integrative processes interplay with each other making information manipulation becomes part of the integrative bargaining process. Integrative negotiation contrasts with distributive bargaining in which the parties are trying to distribute a fixed resource, and where if an agent wins another agent looses. Distributive negotiation predicts that one party can only gain at the other party’s expense. The key characteristics that distinguish integrative negotiations from distributive ones are: creation of value; focus on interests and not positions; openness and exchange of relevant O pe n A cc es s D at ab as e w w w .in te ch w eb .o rg