An ASP-Based Approach to Counterfactual Explanations for Classification
This work addresses the need for interpretable AI by providing a method for counterfactual explanations, but it appears incremental as it builds on existing answer-set programming approaches.
The paper tackles the problem of generating causality-based counterfactual explanations for classification models by proposing answer-set programs, which can be applied to black-box and logic-based models, with a focus on maximum responsibility explanations and the use of semantic knowledge.
We propose answer-set programs that specify and compute counterfactual interventions as a basis for causality-based explanations to decisions produced by classification models. They can be applied with black-box models and models that can be specified as logic programs, such as rule-based classifiers. The main focus in on the specification and computation of maximum responsibility causal explanations. The use of additional semantic knowledge is investigated.