GROVE: A computationally grounded model for rational intention revision in BDI agents

Leask, Sam (2021) GROVE: A computationally grounded model for rational intention revision in BDI agents. PhD thesis, University of Nottingham.

[img]
Preview
PDF (Thesis - as examined) - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader
Available under Licence Creative Commons Attribution.
Download (1MB) | Preview

Abstract

A fundamental aspect of Belief-Desire-Intention (BDI) agents is intention revision. Agents revise their intentions in order to maintain consistency between their intentions and beliefs, and consistency between intentions. A rational agent must also account for the optimality of their intentions in the case of revision. To that end I present GROVE, a model of rational intention revision for BDI agents. The semantics of a GROVE agent is defined in terms of constraints and preferences on possible future executions of an agent’s plans. I show that GROVE is weakly rational in the sense of Grant et al. and imposes more constraints on executions than the operational semantics for goal lifecycles proposed by Harland et al. As it may not be computationally feasible to consider all possible future executions, I propose a bounded version of GROVE that samples the set of future executions, and state conditions under which bounded GROVE commits to a rational execution.

Item Type: Thesis (University of Nottingham only) (PhD)
Supervisors: Logan, Brian
Alechina, Natasha
Keywords: Belief-Desire-Intention, Agents, Rationality, Artificial Intelligence, Goals, Deliberation, Intention Revision.
Subjects: Q Science > QA Mathematics > QA 75 Electronic computers. Computer science
Faculties/Schools: UK Campuses > Faculty of Science > School of Computer Science
Item ID: 65903
Depositing User: Leask, Sam
Date Deposited: 08 Dec 2021 04:40
Last Modified: 08 Dec 2021 04:40
URI: http://eprints.nottingham.ac.uk/id/eprint/65903

Actions (Archive Staff Only)

Edit View Edit View