分布式数据库的精简频繁模式集及其挖掘算法  

Reducing and mining of frequent patterns in distributed database

作　　者：贾泂[1] 刘群[1] 姜晗[2] 

机构地区：[1]浙江师范大学数理与信息工程学院,浙江金华321004 [2]济宁职业技术学院计算机工程系,山东济宁272000

出　　处：《浙江师范大学学报（自然科学版）》2010年第2期132-137,共6页Journal of Zhejiang Normal University:Natural Sciences

基　　金：国家自然科学基金资助项目(60625204)

摘　　要：对分布式数据库关联规则的挖掘与精简表示进行了研究,把频繁闭项集和最大频繁项集的概念推广到分布式数据库中,提出了在分布式环境下的频繁模式集的精简表示方法,以及一种基于各站点的全局大项目集的全局频繁闭项集和全局最大频繁项集的挖掘算法.该算法具有占用空间少、通信量小等特点.最后用实例进行了证明.Based on thorough investigation of mining and simplifying association rules on distributed databases,the concept of closed frequent itemsets and maximal frequent itemsets was extended to combining with distributed databases.Furthermore,it was proposed a new simplification technique of frequent patterns in distributed environment,together with a fresh mining algorithm of globally closed frequent patterns itemsets and globally maximal frequent itemsets,which was based on local-global itemsets at each distributed station.The characteristics of the proposed methods costed less memory and with less traffic load.The experiments demonstrated the competence of the new algorithms.

关 键 词：分布式数据库 关联规则的精简 频繁闭项集 最大频繁项集 全局最大频繁项集 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]