Using iterative narrowing to enable multi-party negotiations with multiple interdependent issues

ABSTRACT

Multi-issue negotiations are a central part of many coordination challenges, and thus represent an important research topic. Almost all previous work in this area has assumed that negotiation issues are independent, but this is rarely the case in real-world contexts. Our work focuses on negotiation with interdependent issues and, therefore, nonlinear (multi-optimum) agent utility functions. Since the utility functions are typically very complex, the challenge becomes finding high-quality negotiation outcomes without making unrealistic demands concerning how much agents reveal about their utilities. Since negotiations often involve more than two parties, the approach should also be scalable. In this paper, we propose a novel protocol for addressing these challenges, wherein agents approach agreements by iteratively narrowing the space of possible agreements. In the early stages, agents submit rough bids representing promising regions from their utility functions. In later stages, they submit increasingly narrow bids for the subset of those regions that the negotiating parties all liked. We show that our method outperforms existing methods in large nonlinear utility spaces, and is computationally feasible for negotiations with as many as ten agents.

REFERENCES

• 1: Raymond T. Ng , Jiawei Han, CLARANS: A Method for Clustering Objects for Spatial Data Mining, IEEE Transactions on Knowledge and Data Engineering, v.14 n.5, p.1003-1016, September 2002 [doi>10.1109/TKDE.2002.1033770]
• 2: Stuart J. Russell , Peter Norvig, Artificial intelligence: a modern approach, Prentice-Hall, Inc., Upper Saddle River, NJ, 1995

INDEX TERMS

Primary Classification:
  I. Computing Methodologies
  I.2 ARTIFICIAL INTELLIGENCE
        I.2.11 Distributed Artificial Intelligence

General Terms:
Algorithms

Keywords:
multi-issue negotiation