Revising Institutions Governed by Institutions for Compliant Regulations

Conference Paper (2016)
Author(s)

T.C. King (TU Delft - Interactive Intelligence)

Tianzhi Li (Imperial College London)

Marina De Vos (University of Bath)

C.M. Jonker (TU Delft - Interactive Intelligence)

J Padget (University of Bath)

MB Van Riemsdijk (TU Delft - Interactive Intelligence)

Research Group
Interactive Intelligence
DOI related publication
https://doi.org/10.1007/978-3-319-42691-4_11
More Info
expand_more
Publication Year
2016
Language
English
Research Group
Interactive Intelligence
Pages (from-to)
191-208
ISBN (print)
978-3-319-42690-7
ISBN (electronic)
978-3-319-42691-4

Abstract

Institutions governing multi-agent systems (MASs) are a pervasive means to guide agents towards the aims of the MAS (e.g. collecting data) with regulations on the outcomes of agents’ behaviour. Yet, wider organisations/governments often intend to guide the design of institutions governing MAS in meeting different aims (e.g. preserving the rights of agents). A pervasive means to guide the design of MAS-governing institutions (or any institution, for that matter) is to use institutions at higher tiers of governance (e.g. directives, constitutions) to regulate the regulations of institutions at lower tiers of governance (e.g. national legislation, software policies). A recent innovation has been an automated means to determine the compliance of a lower-tier institution’s regulations with a higher-tier’s. However, for a designer of a non-compliant institution there remains a dilemma: be punished for non-compliant regulations or arduously determine and rectify the underlying causes of non-compliance. In this paper we propose a way to automatically determine how to revise an institution to be compliant that also minimises the change in the regulations’ outcomes thus keeping as closely as possible to the institution designers’ original intentions.

No files available

Metadata only record. There are no files for this record.