An approach to improving the maintainability of existing rule bases

被引:5
作者
Higa, K
机构
[1] Hong Kong Univ. of Sci. and Technol., Dept. of Info. and Syst. Management, Kowloon, Clear Water Bay
关键词
expert systems; rule structure; rule base maintenance; rule simplification;
D O I
10.1016/0167-9236(96)00015-2
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
A rule base (RE) can be both logically sound (complete and correct) and difficult to maintain often because of its complex rule structure. From examinations of several RBs in textbooks, in expert systems shells, and in research projects, complex rule structures have been found to be a rather common phenomenon. This paper reveals the side-effects of carrying complex rule structures in RBs and then provides formal definitions of complex rule structures. Next, it introduces a procedure to detect and simplify such structures. The proposed rule simplification procedure will improve the maintainability of RBs. Its use is illustrated through sample RBs.
引用
收藏
页码:23 / 31
页数:9
相关论文
共 17 条
[1]  
[Anonymous], 1979, EXPERT SYST MICROELE
[2]  
Chang C. L., 1990, Expert Systems with Applications, V1, P217, DOI 10.1016/0957-4174(90)90003-D
[3]  
ENAND R, 1988, P 3 KNOWL ACQ KNOWL
[4]  
GEISSAMAN JR, 1988, P 4 ANN ART INT ADV
[5]  
HAYESROTH F, 1985, COMMUNICATIONS ACM, V28
[6]  
HIGA K, 1994, BISWPS894 HONG KONG
[7]  
JACOB RJK, 1990, IEEE T KNOWLEDGE DAT, V2
[8]  
LEHRNER PE, 1989, IEEE T SYSTEMS MAN C, V19
[9]  
LELOUCHE R, 1992, P 3 ANN INT ASS KNOW
[10]  
LIANG TP, 1992, MANAGEMENT SCI, V38