GRADUAL INFERENCE RULES IN APPROXIMATE REASONING

被引:136
作者
DUBOIS, D
PRADE, H
机构
[1] Institut de Recherche en Informatique de Toulouse, Université Paul Sabatier, 31062 Toulouse Cedex
关键词
D O I
10.1016/0020-0255(92)90035-7
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Gradual inference rules of the form "The more X is F, the more Y is G" (or of similar forms with "the less" instead of "the more"), which express a progressive change of the degree to which the entity Y satisfies the gradual property G when the degree to which the entity X satisfies the gradual property F is modified, are often encountered in commonsense reasoning. A representation of such rules by means of fuzzy sets is proposed and discussed. This representation turns out to be based on a special implication function already considered in multiple-valued logic. Patterns of reasoning involving gradual inference rules are formalized. Their links with interpolation mechanisms are pointed out.
引用
收藏
页码:103 / 122
页数:20
相关论文
共 27 条
[1]  
BRUXELLES S, 1987, ACTES C COGNITIVA CE, V1, P445
[2]  
BUISSON JC, 1989, JAN ACT CONV INT ART, P43
[3]  
BUISSON JC, 1986, 2ND P INT C ART INT, P705
[4]  
Dubois, 1980, FUZZY SETS FUZZY SYS
[5]   FUZZY LOGICS AND THE GENERALIZED MODUS PONENS REVISITED [J].
DUBOIS, D ;
PRADE, H .
CYBERNETICS AND SYSTEMS, 1984, 15 (3-4) :293-331
[6]  
DUBOIS D, 1982, FUZZY SET POSSIBILIT, P316
[7]  
Dubois D., 1988, POSSIBILITY THEORY
[8]  
DUBOIS D, 1984, STOCHASTICA, V8, P267
[9]  
DUBOIS D, 1989, 3RD P INT FUZZ SYST, P782
[10]  
DUBOIS D, 1984, 1983 P IEEE INT C CY, P300