A Bayesian Approach for Retrieving Relevant

The problem of nding the set of most relevant cases from a given database, with respect to the decision making situation at hand, is frequently encountered in many real-world domains. In the case-based reasoning framework this task is commonly known as the case matching problem. Case matching is an important problem in several commercially signiicant application areas, such as industrial connguration and manufacturing problems. Earlier approaches to the case matching problem typically rely on some distance measure, e.g., the Euclidean distance, although there is no a priori guarantee that such measures really reeect the useful similarities and dissimilarities between the cases. In this paper we introduce a novel approach to the case matching problem based on Bayesian probability theory, and propose a Bayesian case matching measure for scoring the cases with respect to a given decision making situation. The Bayesian case matching score discussed is currently being applied in a real-world manufacturing problem. The empirical results show that when encountered with cases where many of the feature values have been removed, even a relatively small number of remaining values is suucient for retrieving the original case from the case base by using the proposed measure. The experiments also show that the approach is computationally very eecient.