Compromise as a Way to Promote Convention Emergence and to Reduce Social Unfairness in Multi-Agent Systems
Refereed conference paper presented and published in conference proceedings

Times Cited
Altmetrics Information

Other information
AbstractRecently, the study of social conventions has attracted much attention. We notice that different agents may tend to establish different conventions, even though they share common interests in convention emergence. We model such scenarios to be competitive-coordination games. We hypothesize that agents may fail to establish a convention under these scenarios and introducing the option of compromise may help solve this problem. Experimental study confirms this hypothesis. In particular, it is shown that besides convention emergence is promoted, the undesirable social unfairness is also significantly reduced. In addition, we discuss how the reward of coordination via compromise affects convention emergence, social efficiency and unfairness.
All Author(s) ListShuyue Hu, Ho-fung Leung
Name of Conference31st Australasian Joint Conference on Artificial Intelligence
Start Date of Conference11/12/2018
End Date of Conference14/12/2018
Place of ConferenceWellington
Country/Region of ConferenceNew Zealand
Proceedings TitleLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume Number11320
Pages3 - 15
LanguagesEnglish-United Kingdom

Last updated on 2020-12-08 at 03:15