Title
Normative reasoning with an adaptive self-interested agent model based on Markov decision processes
Abstract
Rational self-interested agents, which act so as to achieve the best expected outcome, should violate the norms if the expected rewards obtained with the defections from the norms surpass the expected rewards obtained by being norm-compliant. It means they should estimate the earnings brought about by the violations and the losses caused by their respective reactions. In this paper, we present a rational self-interested agent model that takes into account the possibility of breaking norms. To develop such model, we employ Markov Decision Processes (MDPs). Our approach consists of representing the reactions for norm violations within the MDPs in such a way that the agent is able to reason about how those violations affect her expected utilities and future options. Finally, we perform an experiment in order to establish comparisons between the model presented in this work and its norm-compliant version.
Year
DOI
Venue
2010
10.1007/978-3-642-16952-6_28
IBERAMIA
Keywords
Field
DocType
norm violation,markov decision process,markov decision processes,future option,rational self-interested agent,expected reward,respective reaction,expected outcome,norm-compliant version,normative reasoning,adaptive self-interested agent model,rational self-interested agent model,expected utility,agent architecture
Earnings,Partially observable Markov decision process,Computer science,Markov decision process,Norm (social),Agent architecture,Artificial intelligence,Normative reasoning
Conference
Volume
ISSN
ISBN
6433
0302-9743
3-642-16951-1
Citations 
PageRank 
References 
3
0.38
11
Authors
3
Name
Order
Citations
PageRank
Moser Silva Fagundes1578.65
Holger Billhardt237236.86
Sascha Ossowski31646158.02