Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Advancing Interactive Explainable AI via Belief Change Theory

Antonio Rago, Maria Vanina Martinez

TL;DR

The paper addresses the lack of principled frameworks for interactive XAI by grounding user feedback in belief-change theory, proposing a logic-based representation of classifiers as explanation knowledge bases $K_M = K_d \cup K_e$. It formalises classifiers and explanations with rules, data/explanation separation, and semantic notions like enforcement, consistency, and $\tau$-coherence, enabling principled updates via belief revision operators. It analyzes how core AGM-style postulates translate to interactive XAI across three scenarios S1–S3, suggesting tailored reformulations and practical operators (e.g., partial meet, credibility-limited, and semi-revision) to balance trust, consistency, and minimal change. The work lays a foundation for transparent, auditable, and regulation-ready interactive explanations, and identifies concrete directions for extending the framework to iterative and multi-user feedback. Overall, it provides a formal methodology for updating explanations in response to user input while quantifying the impact on the knowledge base and model fidelity.

Abstract

As AI models become ever more complex and intertwined in humans' daily lives, greater levels of interactivity of explainable AI (XAI) methods are needed. In this paper, we propose the use of belief change theory as a formal foundation for operators that model the incorporation of new information, i.e. user feedback in interactive XAI, to logical representations of data-driven classifiers. We argue that this type of formalisation provides a framework and a methodology to develop interactive explanations in a principled manner, providing warranted behaviour and favouring transparency and accountability of such interactions. Concretely, we first define a novel, logic-based formalism to represent explanatory information shared between humans and machines. We then consider real world scenarios for interactive XAI, with different prioritisations of new and existing knowledge, where our formalism may be instantiated. Finally, we analyse a core set of belief change postulates, discussing their suitability for our real world settings and pointing to particular challenges that may require the relaxation or reinterpretation of some of the theoretical assumptions underlying existing operators.

Advancing Interactive Explainable AI via Belief Change Theory

TL;DR

The paper addresses the lack of principled frameworks for interactive XAI by grounding user feedback in belief-change theory, proposing a logic-based representation of classifiers as explanation knowledge bases . It formalises classifiers and explanations with rules, data/explanation separation, and semantic notions like enforcement, consistency, and -coherence, enabling principled updates via belief revision operators. It analyzes how core AGM-style postulates translate to interactive XAI across three scenarios S1–S3, suggesting tailored reformulations and practical operators (e.g., partial meet, credibility-limited, and semi-revision) to balance trust, consistency, and minimal change. The work lays a foundation for transparent, auditable, and regulation-ready interactive explanations, and identifies concrete directions for extending the framework to iterative and multi-user feedback. Overall, it provides a formal methodology for updating explanations in response to user input while quantifying the impact on the knowledge base and model fidelity.

Abstract

As AI models become ever more complex and intertwined in humans' daily lives, greater levels of interactivity of explainable AI (XAI) methods are needed. In this paper, we propose the use of belief change theory as a formal foundation for operators that model the incorporation of new information, i.e. user feedback in interactive XAI, to logical representations of data-driven classifiers. We argue that this type of formalisation provides a framework and a methodology to develop interactive explanations in a principled manner, providing warranted behaviour and favouring transparency and accountability of such interactions. Concretely, we first define a novel, logic-based formalism to represent explanatory information shared between humans and machines. We then consider real world scenarios for interactive XAI, with different prioritisations of new and existing knowledge, where our formalism may be instantiated. Finally, we analyse a core set of belief change postulates, discussing their suitability for our real world settings and pointing to particular challenges that may require the relaxation or reinterpretation of some of the theoretical assumptions underlying existing operators.
Paper Structure (6 sections, 3 theorems, 6 equations, 2 tables)

This paper contains 6 sections, 3 theorems, 6 equations, 2 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Formalising Classifiers and Explanations
  4. Interactive Explanations
  5. Revision of Explanation KBs
  6. Discussion and Future Work

Key Result

Lemma 1

Given a classifier $\mathcal{M}$, a rule $r$ is coherent with $\mathcal{M}$ iff $\forall \mathbf{x} \in \mathcal{I}_f(body(r)) \cap D$, $\mathcal{M}(\mathbf{x}) \in \mathcal{I}_c(head(r))$.

Theorems & Definitions (18)

  • Definition 1
  • Definition 2
  • Definition 3
  • Definition 4
  • Definition 5
  • Lemma 1
  • Definition 6
  • Definition 7
  • Theorem 1
  • Theorem 1
  • ...and 8 more