Evaluation Metrics for Symbolic Knowledge Extracted from Machine Learning Black Boxes: A Discussion Paper
This is an incremental discussion paper that identifies a critical gap in interpretability for end-users of opaque AI systems, without proposing a new solution.
The paper addresses the lack of quantitative metrics for assessing the readability of symbolic knowledge extracted from machine learning black boxes, highlighting this as an open issue that hinders automatic comparison and development of autotuning algorithms for knowledge extractors.
As opaque decision systems are being increasingly adopted in almost any application field, issues about their lack of transparency and human readability are a concrete concern for end-users. Amongst existing proposals to associate human-interpretable knowledge with accurate predictions provided by opaque models, there are rule extraction techniques, capable of extracting symbolic knowledge out of an opaque model. However, how to assess the level of readability of the extracted knowledge quantitatively is still an open issue. Finding such a metric would be the key, for instance, to enable automatic comparison between a set of different knowledge representations, paving the way for the development of parameter autotuning algorithms for knowledge extractors. In this paper we discuss the need for such a metric as well as the criticalities of readability assessment and evaluation, taking into account the most common knowledge representations while highlighting the most puzzling issues.