Automated Explanation Selection for Scientific Discovery
This work addresses the need for trust in AI systems by improving explanation selection, but it appears incremental as it builds on existing notions with new properties.
The paper tackles the problem of selecting explanations in Explainable Artificial Intelligence (XAI) by proposing a cycle that combines machine learning with automated reasoning, and it presents a taxonomy of explanation selection problems based on insights from sociology and cognitive science.
Automated reasoning is a key technology in the young but rapidly growing field of Explainable Artificial Intelligence (XAI). Explanability helps build trust in artificial intelligence systems beyond their mere predictive accuracy and robustness. In this paper, we propose a cycle of scientific discovery that combines machine learning with automated reasoning for the generation and the selection of explanations. We present a taxonomy of explanation selection problems that draws on insights from sociology and cognitive science. These selection criteria subsume existing notions and extend them with new properties.