A logical foundation for the case-based reasoning cycle
Date of this Version
Case-based reasoning (CBR) has drawn considerable attention in artificial intelligence (AI) fields with many successful applications in systems such as e-commerce and multiagent systems. For the moment, research and development of CBR basically follows the traditional process model of CBR, i.e., the R4 model and problem space model introduced in 1994 and 1996, respectively. However, there has been no logical analysis for this popular CBR model. This article will fill this gap by providing a unified logical foundation for the CBR cycle. The proposed approach is based on an integration of traditional mathematical logic, fuzzy logic, and similarity-based reasoning. At the same time, we examine the CBR cycle from the knowledge-based (KB) viewpoint. The proposed logical approach can facilitate research and development of CBR.
This document has been peer reviewed.