Can You Tell the Difference? Contrastive Explanations for ABox Entailments
This work addresses the need for more interpretable AI in ontology-based systems by providing targeted explanations for users, though it is incremental as it builds on existing explanation methods.
The paper tackles the problem of explaining why an individual is classified as an instance of a concept while another is not in knowledge bases, by introducing contrastive ABox explanations that simultaneously consider both positive and missing entailments. It develops a computational framework, analyzes complexity for various description logics, and implements a method evaluated on generated problems for realistic knowledge bases.
We introduce the notion of contrastive ABox explanations to answer questions of the type "Why is a an instance of C, but b is not?". While there are various approaches for explaining positive entailments (why is C(a) entailed by the knowledge base) as well as missing entailments (why is C(b) not entailed) in isolation, contrastive explanations consider both at the same time, which allows them to focus on the relevant commonalities and differences between a and b. We develop an appropriate notion of contrastive explanations for the special case of ABox reasoning with description logic ontologies, and analyze the computational complexity for different variants under different optimality criteria, considering lightweight as well as more expressive description logics. We implemented a first method for computing one variant of contrastive explanations, and evaluated it on generated problems for realistic knowledge bases.