Rules of encounter: designing conventions for automated negotiation among computers
Rules of encounter: designing conventions for automated negotiation among computers
Local search algorithms on the Stable Marriage Problem: Experimental Studies
Proceedings of the 2010 conference on ECAI 2010: 19th European Conference on Artificial Intelligence
| Hi-index | 0.00 |
This paper promotes a multiagent approach for solving a well-known problem: the stable marriage problem. The aim is to find a "suitable" mapping between two sets of agents, respecting the individuals' preferences on the possible partners. According to this approach, a solution is the outcome of the negotiations between agents. A straightforward approach consists in agentifying the seminal Gale-Shapley algorithm by distinguishing two agent behaviors: the proposers and the responders. The obtained assignment is unfair: the community of proposers is favored. In this paper, we propose the method Swing which applies the minimal concession strategy for negotiation. Each agent alternatively plays these roles in many bilateral negotiations. Our approach may lead to the emergence of stable matchings which cannot be reached by the Gale-Shapley algorithm. They are more equitable between the two communities, more optimal, and then more acceptable for the whole community.