Rules of encounter: designing conventions for automated negotiation among computers
Rules of encounter: designing conventions for automated negotiation among computers
Multiagent systems and societies of agents
Multiagent systems
Distributed rational decision making
Multiagent systems
Optimal agendas for multi-issue negotiation
AAMAS '03 Proceedings of the second international joint conference on Autonomous agents and multiagent systems
| Hi-index | 0.00 |
A mechanism is presented that allows agents in multiagent systems with private knowledge about resource endowments to establish negotiation groups with commonly known negotiation spaces. This is done by exchanging information. Specific agents that are provided by the system mediate the communication and take over some very complex calculations. The negotiating agents reveal only information to the mediators that they consider relevant for their own objectives or for the current negotiation. A broadcast of all the agents' private knowledge is not required. We label this interactive mechanism social formation of negotiation spaces and groups. Furthermore we drop the often-made assumption of isolated negotiations. Our agents remember other agents that have contributed to reach a beneficial agreement. These agents are assumed to have credits. In later negotiations our agents will be more generous towards agents with credits. The applied credit mechanism does not require that all agents participate nor does it require that other agents are aware of this mechanism. We conclude this report with hypotheses that will be evaluated by simulations.