一种改进的关联规则提取算法  

An Improved Algorithm for Mining Association Rules

作　　者：刘军[1] 谢康林 

机构地区：[1]上海交通大学计算机科学与工程系,上海200030

出　　处：《小型微型计算机系统》2003年第7期1343-1345,共3页Journal of Chinese Computer Systems

摘　　要：运用图论中的完全图知识 ,对关联规则提取过程的第一阶段进行改造 ,把大项集计算转换为集合的并和交两种基本运算 ,并利用候选大项集生成过程中的中间结果对已知大项集进行过滤 ,大大减少不必要的重复计算 。Based on theory of complete graph, an improved algorithm for mining association rules is given in this paper. Large item computing, the first phase of Apriori Algorithm, is realized just by basic computing-the union and minus of set. And at the same time, to improve the speed of generation of large itemsets, several efficient ways are introduced to filter the meta result of large items.

关 键 词：数据挖掘 关联规则 大项集 完全图 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]