挖掘多级概念上的广义关联规则  

Mining Generalized Association Rules on Multilevel Concept

作　　者：高丽平[1] 夏敏捷[1] 

机构地区：[1]中原工学院计算机科学系,河南郑州450007

出　　处：《微机发展》2004年第3期104-106,共3页Microcomputer Development

摘　　要：挖掘关联规则的任务是在给定交易集中,每个交易包含一个数据项集,关联发现函数作用在交易集上,返回各数据项集间存在的关系。现实世界中,普遍存在"多级"的概念,在许多应用中,数据项集之间有用的关联规则常常出现在相对较高的概念层中,但在较低概念层往往可以发现较特殊和专门的信息。文中给出了一个在多级概念层上交互挖掘关联规则的算法,并进行了讨论,实验结果表明此算法提高了数据挖掘的效率和速度,并减少了对系统资源的利用。Multiple layer is a concept that exists in many fields of research.One object may not only belong to defferent higher layers,but derives various lower layers.Useful association rules among data items usually occur in rather highlayers,however,special valuable information can be ofter focus in rather low conceptual layers.This article gives an algorithm on mutual mining the association rules among multiple conceptual layers and come issues about it. The algorithm is turned out to be much more effective and rapid, when running, it use up little system resource.

关 键 词：数据库 数据挖掘 数据仓库 广义关联规则 知识发现 数据项 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]