基于一次性数据库访问策略的关联规则挖掘算法的研究  

The Research on the Mining Association Rules Algorithm Based on the Strategy Accessing to Database Once

作　　者：刘培奇[1] 卢麟[1] 廖福燕[1] 宋阳[1] 

机构地区：[1]西安建筑科技大学信息与控制工程学院,陕西西安710055

出　　处：《微电子学与计算机》2010年第12期22-25,共4页Microelectronics & Computer

基　　金：国家自然科学基金项目(60673170);陕西省教育厅自然科学基金专项(08JK318);西安建筑科技大学校人才基金项目(RC0618)

摘　　要：针对大型数据库挖掘中需要多次访问数据库和效率较低的局限性,提出一次性数据库访问策略,设计了基于Apriori的Apriori_ADO算法.经过实验分析,Apriori_ADO算法降低了算法的时间和空间开销,提高了关联规则挖掘效率.算法具有很强的实用性,已用于超市中顾客消费知识的挖掘,并取得了满意的结果.This article analyzes the Apriori algorithm,and points out the limitations that it requires many accesses in the large-scale database mining and its efficiency is low.Based on the strategy accessing to DB once,it designs Apriori_ADO algorithm based on the Apriori.This algorithm avoids cutting frequently.Experimental analysis shows that the new improved Apriori_ADO algorithm reduces the time and space,and improves the mining efficiency of association rules.Apriori_ADO algorithm has strong practicality;it has been used for supermarkets in customer consumer knowledge mining,and has achieved satisfying results.

关 键 词：关联规则 APRIORI算法 频繁项集 支持度 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]