Can Automated Agents Proficiently Negotiate with Humans?
The article discusses research into the design of computer programs which can conduct negotiations. An overview of the challenges involved in designing a negotiator advanced enough to be capable of negotiating successfully with humans is presented. Examples of various programs which have achieved so...
| Publicado en: | Communications of the ACM Vol. 53; no. 1; pp. 78 - 89 |
|---|---|
| Autores principales: | , |
| Formato: | Artículo |
| Publicado: |
Association for Computing Machinery
Jan2010
|
| Materias: | |
| Acceso en línea: | Ver este registro en EBSCOhost |
| fields | @attributes: recordID: 1 pdfLink: plink: https://search.ebscohost.com/login.aspx?direct=true&db=hlh&AN=47220616&site=ehost-live header: @attributes: shortDbName: hlh uiTerm: 47220616 longDbName: Humanities International Complete uiTag: AN controlInfo: bkinfo: jinfo: jid: 00010782 ACM jtl: Communications of the ACM issn: 00010782 maglogo: N pubinfo: dt: Jan2010 vid: 53 iid: 1 pid: 68 pub: Association for Computing Machinery artinfo: ui: 47220616 10.1145/1629175.1629199 ppf: 78 ppct: 11 formats: tig: atl: Can Automated Agents Proficiently Negotiate with Humans? aug: au: RAZ LIN KRAUS, SARIT affil: Postdoctoral Fellow, computer science department, Bar-Ilan University, Ramat-Gan, Israel. Professor of computer science department, Bar-Ilan University, Ramat-Gan, Israel. Adjunct professor, Institute for Advanced Computer Studies, University of Maryland, College Park, MD. su: Negotiation Computer programming Artificial intelligence Game theory Simulation games Decision theory sug: subj: Negotiation Computer programming Artificial intelligence Game theory Simulation games Decision theory ab: The article discusses research into the design of computer programs which can conduct negotiations. An overview of the challenges involved in designing a negotiator advanced enough to be capable of negotiating successfully with humans is presented. Examples of various programs which have achieved some of the necessary benchmarks are discussed, such as Diplomat, Colored-Trails, and QOAgent. The strategies used by the designers of these programs are analyzed. These include randomization to simulate non-deterministic behavior, a mechanism for making concessions, and a database to store acquired information. pubtype: Periodical doctype: Article src: R language: English refInfo: copyright: @attributes: flag: Y dt: @attributes: year: 2010 holdings: @attributes: islocal: N |
|---|