Designing Flexible Negotiation Agent with Relaxed Decision Rules

This article presents a model for designing negotiation agent with two distinguishing features: 1) a marketdriven strategy and 2) a set of relaxed decision rules. Market-driven agents make adjustable rates of concession by reacting to changing market situations. In determining the amount of concession for each negotiation round, a market-driven agent is guided by three mathematical functions of remaining trading time, trading opportunity, and competition. Furthermore, agents in this research are also guided by a set of fuzzy rules for determining whether to relax their trading aspirations in the face of (very) high negotiation pressures such as fast approaching deadlines, and high competitions. Empirical results suggest that on average, when compared to Sim’s market-driven, agents in this research achieved 1) higher success rates in reaching a deal and 2) higher expected utilities.

• Type: Conference
• Year: 2003
• Where: IAT
• Authors: Kwang Mong Sim, Shi Yu Wang
• Keywords: IAT 2003, Intelligent Agent, Market-driven Agent, Negotiation, Relaxed Decision Rules