一种挖掘最小蕴涵规则集的通用算法  

A GENERAL ALGORITHM FOR MINING IMPLICATION RULES

作　　者：时百胜[1] 刘宗田[2] 余泓[1] 

机构地区：[1]苏州科技学院应用数学系,江苏苏州215009 [2]上海大学计算机工程与科学学院,上海200072

出　　处：《计算机应用与软件》2007年第9期187-188,193,共3页Computer Applications and Software

摘　　要：给出一种挖掘最小蕴涵规则集的通用算法,该算法基于闭包运算,利用闭包运算产生所有闭集,从闭集格中导出最大∧-不可约集及其补集,然后产生最小蕴涵规则集;给出了计算所有闭集的新算法,同Ganter算法相比,该算法充分考虑闭包运算的特性,使得冗余计算显著减少,提高了算法效率。A general algorithm is proposed for mining minimal implication rules from large database. It is based on closure operation. By means of closure operation, the family of closed sets is computed. From closed sets, maximal meet-irreducible sets and its complements are derived ,and all minimal non-trivial implication rules can be generated. If closure operation on a set is defined,this algorithm can all implication rules of this set.

关 键 词：闭集格 闭包运算 最大∧-不可约集 最小蕴涵规则集 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]