A Dynamic Preference Logic for reasoning about Agent Programming

TL;DR

This paper introduces a dynamic preference logic framework to model and reason about BDI agent mental attitudes, enabling formal analysis and specification of agent programming and rationality properties.

Contribution

It develops a novel logical approach connecting dynamic preference logic with BDI agent programming, including a new interpretation of preferences as priority graphs.

Findings

Provides a formal encoding of BDI attitudes using dynamic preference logic

Shows how to interpret BDI programs within this logical framework

Demonstrates the logic's use in specifying and verifying rationality properties

Abstract

In this work, we investigate the use of Dynamic Preference Logic to encode BDI mental attitudes. Further, exploring this codification and the representation of preferences over possible worlds by preferences over propositional formulas, here called priority graphs, we comment on how to interpret BDI agent programs in this logic. Also, using the connection between dynamic operations defined over preference models and their encoding as transformations on priority graphs, we show how our logic can be used not only to reason about agent programs, but as a tool to specify reasoning mechanisms to guarantee certain properties in the theory of rationality for the programming language.