Case-based Reasoning (CBR) means reasoning from prior examples. It involves retaining a memory of previous problems and their solutions and, by referencing this knowledge, new problems are compared, and the previous successful solutions are adapted and applied to the new problem situation. Therefore, the effectiveness of CBR systems largely depends on the quality of the past cases and the maintenance of its reasoning ability. Broadly speaking, case-base maintenance (CBM) activities can be divided into two types: (1) qualitative maintenance, and (2) quantitative maintenance. Qualitative maintenance refers to the assurance of the correctness, consistency and completeness of a CBR system, while quantitative maintenance refers to the problem solving efficiency of a CBR system. Experience with the growing number of deployed CBR systems has led to the awareness of their maintenance. Consequently, understanding and developing good practical CBM strategies are crucial to sustaining and improving the acceptance of CBR systems. In this chapter, the background and basic concepts of CBM will be reviewed, and followed by a brief discussion of the current research work on CBM. Next, the use of fuzzy set, rough set and fuzzy integral for maintaining a distributed case-base reasoning system is demonstrated, along with some experimental testing using an example case-base in the travel domain.
|Title of host publication||Soft computing approach to pattern recognition and image processing|
|Number of pages||30|
|ISBN (Print)||9789812382511, 9812382518|
|Publication status||Published - 2002|