Unsupervised Expressive Rules Provide Explainability and Assist Human Experts Grasping New Domains
This assists domain experts in grasping new domains by providing explainable, unsupervised rules, though it is incremental as it builds on existing clustering and explainability methods.
The paper tackles the problem of domain experts lacking labeled data and unsatisfactory domain adaptation when approaching new corpora, presenting an unsupervised method that reveals human-readable rules to cluster data by prominent categories, with evaluation showing usefulness in identifying target categories and interpretability in a user study.
Approaching new data can be quite deterrent; you do not know how your categories of interest are realized in it, commonly, there is no labeled data at hand, and the performance of domain adaptation methods is unsatisfactory. Aiming to assist domain experts in their first steps into a new task over a new corpus, we present an unsupervised approach to reveal complex rules which cluster the unexplored corpus by its prominent categories (or facets). These rules are human-readable, thus providing an important ingredient which has become in short supply lately - explainability. Each rule provides an explanation for the commonality of all the texts it clusters together. We present an extensive evaluation of the usefulness of these rules in identifying target categories, as well as a user study which assesses their interpretability.