基于多个最小支持度的频繁项目集挖掘算法  

Frequent Itemset Mining Algorithm Based on Multiple Minimum Support Degrees

作　　者：陈福集[1] 李福平[1] 

机构地区：[1]福州大学公共管理学院,福州350108

出　　处：《计算机工程》2011年第24期36-37,41,共3页Computer Engineering

基　　金：国家杰出青年科学基金资助项目(70925004)

摘　　要：传统的关联规则挖掘算法不能在同一事务数据库中连续挖掘多个最小支持度的频繁项目集。为此,提出基于多个最小支持度的频繁项目集挖掘算法。运用集合论定义模型库的概念,将事务数据库转化成模型库,通过检索模型库得到频繁项目集,从而降低频繁项目集的挖掘时间。实验结果表明,该算法的挖掘效率高于Apriori算法。To the demand of a continuous mining frequent itemset in the same transaction database under multiple minimum support degree, this paper proposes frequent itemset mining algorithm based on multiple minimum support degrees. The algorithm uses set theory, leads into model library, converts the transaction database into a model library, and searches model library to obtain frequent itemset. The algorithm reduces the time of frequent itemset mining and improves efficiency of frequent itemset mining. Experimental results show this algorithm is more efficient than Apriori algorithm.

关 键 词：关联规则 数据挖掘 最小支持度 模型库 频繁项目集 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]