Artificial intelligence: a modern approach
Artificial intelligence: a modern approach
Communications of the ACM
Composition of Secure Multi-Party Protocols: A Comprehensive Study
Composition of Secure Multi-Party Protocols: A Comprehensive Study
Modeling complex multi-issue negotiations using utility graphs
Proceedings of the fourth international joint conference on Autonomous agents and multiagent systems
Human vs. Computer Behaviour in Multi-Issue Negotiation
RRS '05 Proceedings of the Rational, Robust, and Secure Negotiation Mechanisms in Multi-Agent Systems (RRS'05) on Multi-Agent Systems
Approximate and online multi-issue negotiation
Proceedings of the 6th international joint conference on Autonomous agents and multiagent systems
A preliminary result on a representative-based multi-round protocol for multi-issue negotiations
Proceedings of the 7th international joint conference on Autonomous agents and multiagent systems - Volume 3
Analysis of privacy loss in distributed constraint optimization
AAAI'06 Proceedings of the 21st national conference on Artificial intelligence - Volume 1
Reaching envy-free states in distributed negotiation settings
IJCAI'07 Proceedings of the 20th international joint conference on Artifical intelligence
Multi-issue negotiation protocol for agents: exploring nonlinear utility spaces
IJCAI'07 Proceedings of the 20th international joint conference on Artifical intelligence
Hi-index | 0.00 |
We envision the future in which large number of participants collaborate, negotiate, and reach consensus through computer supported negotiation support system against a global problem in the world. Multi interdependent issues negotiation has been studied widely since such most real-world negotiation involves multiple interdependent issues. Our work focuses on negotiation with multiple interdependent issues, in which agent utility functions are nonlinear. Existing works have not yet concerned with agents' private information that should be concealed from others in negotiations. In this paper, we propose Distributed Mediator Protocol (DMP) to securely find the agreements that satisfy the Pareto optimality. DMP successfully conceals agents' private information. Then, we propose the following two measures for selecting the final agreements from the set of Pareto-optimal contracts. First, we propose "approximated fairness," which represents how fair the contract is for each agent. We employ deviation for measuring the difference of utilities achieved by agents. Second, we propose the rate of Nash bargaining solution, which represents how close the contract is to the Nash bargaining solution. The Nash bargaining solution maximizes the product of each agent's utilities (Nash product) in our model. The approximated fairness helps the mediator find the contract that is close to the Nash bargaining solution. This is because the Nash product will be increased if approximated fairness becomes small. Moreover, we compare DMP with a search algorithm (called Direct Search) to find a contract which maximizes Nash products directly. Direct Search is usually better for finding the Nash bargaining solution. However, Direct Search isn't always better from the Pareto Optimality and Fairness if the situation is difficult for finding the Nash bargaining solution. We compare DMP with Direct Search from Nash product and Pareto optimality in the experiments.