What Makes a Good Explanation?: A Harmonized View of Properties of Explanations

Varshini Subhash, Zixi Chen, Marton Havasi, Weiwei Pan, Finale Doshi-Velez

Keywords: interpretability, explainability, explanation properties, XAI, XML

TL;DR: We survey explanation properties defined in interpretable ML literature, synthesize and unify them based on what they measure, and describe the trade-offs between different formulations of these properties.

Abstract: Interpretability provides a means for humans to verify aspects of machine learning (ML) models. Different tasks require explanations with different properties. However, presently, there is a lack of standardization in assessing properties of explanations: different papers use the same term to mean different quantities, and different terms to mean the same quantity. This lack of standardization prevents us from rigorously comparing explanation systems. In this work, we survey explanation properties defined in the current interpretable ML literature, we synthesize properties based on what they measure, and describe the trade-offs between different formulations of these properties. We provide a unifying framework for comparing properties of interpretable ML.