Grant-in-Aid for Scientific Research (C).
|Research Institution||Nagoya Institute of Technology|
SHINTANI Toramatsu Nagoya Institute of Technology, Dept. of Intelligence and Computer Science, Professor, 工学部, 教授 (00252312)
|Project Fiscal Year
1998 – 2001
Completed(Fiscal Year 2001)
|Budget Amount *help
¥3,100,000 (Direct Cost : ¥3,100,000)
Fiscal Year 2001 : ¥900,000 (Direct Cost : ¥900,000)
Fiscal Year 2000 : ¥700,000 (Direct Cost : ¥700,000)
Fiscal Year 1999 : ¥600,000 (Direct Cost : ¥600,000)
Fiscal Year 1998 : ¥900,000 (Direct Cost : ¥900,000)
|Keywords||agent / cooperation model / utility theory / persuasion protocol / mobile agent / decision support / group decision making / multiple negotiation / エージェント / 協調モデル / 効用理論 / 説得プロトコル / モバイルエージェント / 意思決定支援 / グループ意思決定 / 多重交渉 / 意志決定支援 / グループ意志決定 / 電子商取引支援 / スケジューリング / 合意形成アルゴリズム / 分散型意思決定支援 / 制約解消アルゴリズム|
We studied on a coalition formation algorithm based on a multiagent system for distributed decision support systems. We propose a new multiagent negotiation method that includes the multiple negotiations and the preference revision in persuasion among agents. We have implemented some application systems to see how effectively the negotiation method can be used as follows.
(a) An autonomous bidding strategy planning system in online auctions
We developed an algorithm for agents that can autonomously make an effective bidding strategy in online auctions. Our algorithm can also make an effective strategy in difficult cases, like, multiple items have complex relations (e.g. time and preference orders). We developed online auction support systems using the algorithm, which support bidding and browsing information.
(b) A coalition formation system using argumentation-based negotiation among agents
We developed a coalition formation algorithm using argumentation-based negotiation among agents. We
formalized a negotiation model for multiagent negotiation that enables agents to reach a reasonable and preferable agreement among users and developed a mechanism based on the model that can resolve a reasonable agreement among agents.
c A distributed meeting scheduler using a multiple negotiation
We developed a distributed meeting scheduler using a multiple negotiation. In meeting scheduling system, an agent is assigned to a user who plans private and public schedule. An agent negotiates with other agents about making a public schedule by referring the user's private schedule and preferences. The multiple negotiations facilitates reaching an agreement among agents effectively. The result shows that the multiagent negotiation based on private preferences is an effective method for a distributed meeting scheduler.
The research results can be summarized as follows; (1) we propose new building methodologies for mobile agents and their applications. Mobile agents are promising and advanced technology to utilize world-wide information networks. (2) We developed methods that can be used to build intelligent electronic commerce support systems. Our methods will be technical seeds for promoting future social information infrastructures. (3) We developed the coalition formation algorithm among agents for multiagent-mediated distributed decision support system. This algorithm can be applied to real distributed decision support systems widely. Less