Dubois et al. have already published work on the use of fuzzy reasoning in case-based decision support. This paper builds on that prior work, presenting a theoretical formulation for, and an example usage of, two ways of using fuzzy reasoning in case-based decision support.
The target applications are recommender systems; the paper presents examples, but no specific empirical evaluation. The first method discussed is a fuzzy set-based counterpart to earlier work on case-based decision theory [1]. The second method is a fuzzy set version of earlier work on constraint-based nearest-neighbor classification [2]. The paper is reasonably self-contained; it includes a brief account of the basics of fuzzy sets, an overview of applications of fuzzy sets in case-based decision support, and introductions to the two mentioned theories that are being built upon.
The work is clearly presented, and includes useful commentary presenting the authors’ views of the benefits of fuzzy reasoning. The paper will be accessible to a reader with a sound grasp of formal methods and a general sense of the ideas behind recommender systems, but with no particular prior exposure to fuzzy sets or methods.