Adapting to agents' personalities in negotiation

Shavit Talman, Meirav Hadad, Yakov Gal, Sarit Kraus

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract / Description of output

To establish cooperative relationships, agents must be willing to engage in helpful behavior and to keep their commitments to other agents. However, in uncertain and dynamic environments, it is difficult to identify the degree of helpfulness of other agents. This paper describes a model in which agents' helpfulness is characterized in terms of cooperation and reliability. An agent chooses an action based on its estimate of others' degree of helpfulness given the dependency relationships that hold between the agent and others. This model was evaluated in a negotiation game in which players needed to exchange resources to reach their goals, but did not have information about each others' resources. Results showed that agents using the model could identify and adapt to others' varying degree of helpfulness even while the other agents were constantly changing their strategy. Moreover, agents that varied their degree of helpfulness depending on their estimate of others' helpfulness outperformed agents who did not, as well as increased the social welfare of the group.
Original languageEnglish
Title of host publicationProceedings of the fourth international joint conference on Autonomous agents and multiagent systems
Place of PublicationThe Netherlands
PublisherACM
Pages383-389
Number of pages7
ISBN (Print)1-59593-093-0
DOIs
Publication statusPublished - 25 Jul 2005
EventFourth international joint conference on Autonomous agents and multiagent systems - , Netherlands
Duration: 25 Jul 200529 Jul 2005

Conference

ConferenceFourth international joint conference on Autonomous agents and multiagent systems
Abbreviated titleAAMAS '05
Country/TerritoryNetherlands
Period25/07/0529/07/05

Fingerprint

Dive into the research topics of 'Adapting to agents' personalities in negotiation'. Together they form a unique fingerprint.

Cite this