A Survey on Retrieval Strategy for Case-Based Reasoning using Modified - USIMSCAR for Object Oriented Cases

Abstract

Case based reasoning is the problem solving technique which is fully based on the solutions of similar past problems. In CBR system, retrieval is an important part since the overall performance is based on it. Aim of this paper is to retrieve the combined set of each cases and rules relevant to the target drawback, using the combined knowledge of similarity and association. This paper proposes the retrieval strategy in CBR System for object oriented cases using modified - USIMSCAR algorithm. It also argues the performance of the generated association rules using SCAR algorithm in CBR systems. Association rules are created using ARM techniques and OO-datasets are used for evaluation. Modified - USIMSCAR algorithm is applied in medical domain.

Keywords: CBR, Association knowledge (AK), Association rule mining (ARM), Case-Based Reasoning, Multilevel association rules, Modified-USIMSCAR(unified similarity and soft- matching class association rule)