Authors: Mu-Jung Huang; Mu-Yen Chen; Meng-Yu Lin
Addresses: Department of Accounting, National Changhua University of Education, No. 2, Shi-Da Road, Changhua, 50058, Taiwan ' Department of Information Management, National Taichung University of Science and Technology, No. 129, Sec. 3, San-Min Road, Taichung City, 40444, Taiwan ' Department of Information Management, National Changhua University of Education, No. 2, Shi-Da Road, Changhua, 50058, Taiwan
Abstract: Most current case-based reasoning (CBR) systems are designed for solving single-condition problems. However, there are multiple-condition problems (i.e., those involving several different problem conditions, such as personal computer [PC] problems) that also need specific solutions. Thus, this paper attempts to integrate CBR concepts with knowledge mapping in a CBR system designed to solve multiple-condition problems. This study makes three critical contributions: 1) it adopts a user-oriented approach to measuring case similarity and allows the user to select from a list of category features to characterise the new problem; 2) it presents a knowledge mapping algorithm suitable for applying the knowledge of experts to the proposed problems; 3) it uses a prototype to demonstrate how smoothly the proposed approach can solve the target PC problems. Our results indicate that a CBR system with a knowledge mapping approach is suitable for solving multiple-condition problems.
Keywords: case-based reasoning; knowledge mapping; multiple-condition problem solving.
International Journal of Social and Humanistic Computing, 2017 Vol.2 No.3/4, pp.187 - 202
Available online: 19 Jun 2017 *Full-text access for editors Access for subscribers Purchase this article Comment on this article