Extraction of domain knowledge from databases based on rough set theory

Automated knowledge acquisition is an important research area to solve the bottleneck problem in developing expert systems. For this purpose, there have been proposed several methods of inductive learning, such as induction of decision trees, AQ method, and neural networks. However, most of the approaches focus on inducing rules which classify cases correctly. On the contrary, medical experts also learn other information which is important for medical diagnostic procedures from databases. In this paper, a rule-induction system, called PRIMEROSES (Probabilistic Rule Induction Method based on Rough Sets version 3.0), is introduced. This program first analyzes the statistical characteristics of attribute-value pairs from training samples, then determines what kind of diagnosing model can be applied to these training samples. Finally it extracts domain knowledge needed for other diagnostic procedures, based on a selected diagnosing model. PRIMEROSES is evaluated on clinical databases on headache, and the induced results are compared with domain knowledge acquired from medical experts. The experimental results show that our proposed method correctly not only selects a diagnosing model, but also extracts domain knowledge.