@inproceedings{e6e975707b594a29bbb685ff880d5226,
title = "Convention Emergence with Congested Resources",
abstract = "Norms and conventions enable coordination in populations of agents by establishing patterns of behaviour, which can emerge as agents interact with their environment and each other. Previous research on norm emergence typically considers pairwise interactions, where agents{\textquoteright} rewards are endogenously determined. In many real-life domains, however, individuals do not interact with one other directly, but with their environment, and the resources associated with actions are often congested. Thus, agents{\textquoteright} rewards are exogenously determined as a function of others{\textquoteright} actions and the environment. In this paper, we propose a framework to represent this setting by: (i) introducing congested actions; and (ii) adding a central authority, that is able to manipulate agents{\textquoteright} rewards. Agents are heterogeneous in terms of their reward functions, and learn over time, enabling norms to emerge. We illustrate the framework using transport modality choice as a simple scenario, and investigate the effect of representative manipulations on the emergent norms.",
keywords = "Congestion games, Conventions, Norm emergence",
author = "Priel Levy and Nathan Griffiths",
note = "Publisher Copyright: {\textcopyright} 2021, Springer Nature Switzerland AG.; 18th European Conference on Multi-Agent Systems, EUMAS 2021 ; Conference date: 28-06-2021 Through 29-06-2021",
year = "2021",
doi = "10.1007/978-3-030-82254-5_8",
language = "الإنجليزيّة",
isbn = "9783030822538",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Science and Business Media Deutschland GmbH",
pages = "126--143",
editor = "Ariel Rosenfeld and Nimrod Talmon",
booktitle = "Multi-Agent Systems - 18th European Conference, EUMAS 2021, Revised Selected Papers",
address = "ألمانيا",
}