Negotiating with Unknown Opponents Toward Multi-lateral Agreement in Real-Time Domains

Automated negotiation has been gained a mass of attention mainly because of its broad application potential in many fields. This work studies a prominent class of automated negotiations – multi-lateral multi-issue negotiations under real-time constraints, where the negotiation agents are given no prior information about their opponents’ preferences over the negotiation outcome space. A novel negotiation approach is proposed that enables an agent to obtain efficient agreements in this challenging multi-lateral negotiations. The proposed approach achieves that goal by, (1) employing sparse pseudo-input Gaussian processes (SPGPs) to model opponents, (2) learning fuzzy opponent preferences to increase the satisfaction of other parties, and (3) adopting an adaptive decision-making mechanism to handle uncertainty in negotiation.