Introspective revision of structured belief bases with argumentative inference

Abstract

This work addresses the interaction between two significant areas of research in Knowledge Representation and Reasoning: Belief Revision and Argumentation. Both areas focus on determining the validity of the beliefs within an agent’s knowledge base. In belief revision, the emphasis is on modifying beliefs to sustain a coherent knowledge base for the agent. This process generally involves incorporating new information while eliminating existing beliefs. Conversely, argumentation research focuses on assessing the epistemic state of the beliefs within that knowledge base. The decision to accept a belief is made after evaluating all arguments both in favor of and against the claim. Each area contributes tools that model human common sense reasoning as applied ineveryday situations. In recent years, efforts have been made to formalize the interaction between these two areas by applying belief revision strategies to argumentative systems and utilizing formal argumentation techniques to guide change processes in belief revision. One such approach employs stratified belief bases, enhanced with argumentative inference mechanisms. Here, a definition is proposed for the introspective revision of stratified belief bases. The concept involves the introduction of change revision operators that can be applied to astratified belief base. These operators are designed to either eliminate conflicting arguments or modify certain beliefs to accommodate new beliefs, resulting in a change in the status of specific arguments.