Classification-based problem-solving in case-based reasoning

* Final gross prices may vary according to local VAT.

Get Access

Abstract

In this paper, we present a study on the retrieval and adaptation operations for case-based reasoning, in the context of object-based representations. First, the paper describes the case-based reasoning cycle and the associated problem-solving process. Details are given on problem formalization, the organization and representation of cases and case indexes. Indexes are represented as frames lying in a subsumption hierarchy. Therefore, the links between retrieval, adaptation, and classification, are very close. Retrieval and adaptation are analyzed through three main operations, namely complete, incomplete and approximate classification, corresponding to three different ways for handling indexes and cases in object-based representations. The paper ends with a discussion on the topics presented here, and points out future works completing and extending this study.