A Causal Argumentation Method for Explainability of Machine Learning Models

TL;DR

This paper introduces a novel explainability method for machine learning models that combines causality detection with argumentation frameworks to clarify decision-making processes.

Contribution

It integrates causal discovery with bipolar argumentation frameworks to enhance interpretability of model predictions.

Findings

Effectively identifies causal relationships among features.

Provides explanations using argumentation semantics.

Outperforms standard post-hoc explainability methods on benchmarks.

Abstract

Explainable AI (XAI) methods identify which features are relevant to a model's predictions but often fail to clarify why certain decisions are made. In this work, we present a novel method that integrates causality with argument-based reasoning to explain why models may be making predictions. Our approach first identifies causal relationships among variables using causal discovery methods and then translates these into a Bipolar Argumentation Framework (BAF) to represent supportive and opposing interactions among features. By using semi-stable semantics, we find extensions of features that explain why certain outcomes may have been chosen. We demonstrate our method on two benchmark datasets and compare its results against standard post-hoc explainability approaches.