一种挖掘频繁项的新方法  

A New Method of Mining Frequent-Item

作　　者：陈冰[1] 张化祥[1] 

机构地区：[1]山东师范大学信息科学与工程学院,山东济南250014

出　　处：《计算机技术与发展》2008年第8期118-120,125,共4页Computer Technology and Development

基　　金：山东省科技攻关计划(2005GG4210002);山东省青年科学家科研奖励基金(2006BS01020);山东省教育厅科技计划项目(J07YJ04)

摘　　要：介绍了关联规则挖掘的情况,然后对关联规则挖掘算法进行分析,并在此分析的基础上对经典的Apriori算法作出了进一步的改进,从而提出了这种改进的关联规则挖掘算法——Apriori-New算法。Apriori-New算法只需对数据库扫描一次,并在扫描过程中通过不断将被标记为频繁项的项集提取出来,最终找出所有的频繁项集。通过一个简单的实例说明了该算法的扫描过程,从而体现了该Apriori-New算法的效率及其所具有的实用性。Provides a survey of the study in association rule generation. And then makes an analysis of the algorithm of association rule generation. On the basis of the analysis, the classical algorithm of Apriori is analyzed. Meanwhile the algorithm is making a further modification. Then an improved ,Apriori algorithm of Aprior-New is proposed. Due to its advantage of scanning DB only once,during the process of scanning, the frequent - items are marked and selected. In the end, all of the frequentitems can be found. A simple example is used to show the process of scanning. Then the new algorithm - Apriori - New is proposed with high efficiency and certain practical significance.

关 键 词：数据挖掘 关联规则 APRIORI算法 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]