Trepan Reloaded: A Knowledge-driven Approach to Explaining Artificial Neural Networks
This work addresses the need for safer and more trustworthy AI by improving explainability for users in critical domains like finance and medicine, though it is incremental as it builds on an existing method.
The authors tackled the problem of making global post-hoc explanations of neural networks more understandable by incorporating domain knowledge via ontologies into the Trepan algorithm, which generates decision trees. Their user study in finance and medicine domains showed that decision trees produced with their knowledge-driven approach were more understandable than those from standard Trepan, as measured by syntactic complexity, response time, accuracy, user confidence, and reported understandability.
Explainability in Artificial Intelligence has been revived as a topic of active research by the need of conveying safety and trust to users in the `how' and `why' of automated decision-making. Whilst a plethora of approaches have been developed for post-hoc explainability, only a few focus on how to use domain knowledge, and how this influences the understandability of global explanations from the users' perspective. In this paper, we show how ontologies help the understandability of global post-hoc explanations, presented in the form of symbolic models. In particular, we build on Trepan, an algorithm that explains artificial neural networks by means of decision trees, and we extend it to include ontologies modeling domain knowledge in the process of generating explanations. We present the results of a user study that measures the understandability of decision trees using a syntactic complexity measure, and through time and accuracy of responses as well as reported user confidence and understandability. The user study considers domains where explanations are critical, namely, in finance and medicine. The results show that decision trees generated with our algorithm, taking into account domain knowledge, are more understandable than those generated by standard Trepan without the use of ontologies.