Integrating Prior Knowledge in Post-hoc Explanations

TL;DR

This paper introduces a formal framework for incorporating prior knowledge into post-hoc interpretability methods in XAI, exemplified by a new counterfactual explanation method called KICE, which improves explanation relevance.

Contribution

It proposes a general cost function framework to embed prior knowledge into any interpretability method, and develops KICE, a novel counterfactual explanation technique utilizing this framework.

Findings

KICE generates more relevant counterfactuals with prior knowledge.

Experimental results show improved interpretability and personalization.

KICE outperforms reference methods on benchmark datasets.

Abstract

In the field of eXplainable Artificial Intelligence (XAI), post-hoc interpretability methods aim at explaining to a user the predictions of a trained decision model. Integrating prior knowledge into such interpretability methods aims at improving the explanation understandability and allowing for personalised explanations adapted to each user. In this paper, we propose to define a cost function that explicitly integrates prior knowledge into the interpretability objectives: we present a general framework for the optimization problem of post-hoc interpretability methods, and show that user knowledge can thus be integrated to any method by adding a compatibility term in the cost function. We instantiate the proposed formalization in the case of counterfactual explanations and propose a new interpretability method called Knowledge Integration in Counterfactual Explanation (KICE) to optimize it. The paper performs an experimental study on several benchmark data sets to characterize the counterfactual instances generated by KICE, as compared to reference methods.