摘要
本文提出了一种基于用户指导的多关系关联规则挖掘算法,借鉴有向图的概念动态的选择最优关键表,并利用元组ID传播的思想使多表间无需物理连接而能直接进行关联规则挖掘,同时引入用户指导的概念,提高了用户的满意程度及海量数据挖掘的效率和精确度.该算法能够直接支持关系数据库,且运行效率远远高于基于ILP技术的多关系关联规则挖掘算法.
A multi-relational association rule mining algorithm with guidance of user is proposed in this paper. The concept of oriented graphic used to dynainic choice most superior key table, and a tuple ID propagation approach is used to solve directly the association rule mining problem with multiple database relations, and the concept of user's guidance is introduced. The usage of this approach improves user's satisfaction of the mining result. Compared with the traditional algorithm, it improves the accuracy rate and supports multi-relational database directly, so its running rate is much higher than that of the ILP based multi-relational association rule mining methods.
出处
《微计算机信息》
2009年第24期124-126,共3页
Control & Automation
基金
基金申请人:郭景峰
项目名称:关系数据挖掘理论研究
基金颁发部门:国家自然科学基金委(60673136)
关键词
多关系数据挖掘
多关系关联规则
用户指导
有向图
元组ID传播
multi-relational data mining
multi--relational association rule
user's guidance
Oriented graphic
tuple ID propagation