摘要
给出了一种有效的频繁项双空间挖掘方法,充分利用事务数据库的二元特性,通过双空间映射把数据库的项目维和事务维联系在一起,提高了频繁项集的挖掘效率。计算机实验数据表明,双空间搜索挖掘方法对频繁项的数据挖掘是非常有效的,与传统的Apriori方法相比,新方法对数据扩散率和频繁项长短(最小支持度变化)均不敏感,挖掘效率提高很多。
The paper presents an efficient dual space search algorithm for mining frequents. The algorithm takes full advantage of the duality characteristic in the transaction database via dual space mapping. The items dimension and transaction dimension in the transaction database are combined to increase the efficiency of mining frequents. It proves that dual space search algorithm is very efficient to mining frequents by way of experiment on the computer, new method is not sensitive to scale of data or the length of frequents.
出处
《计算机工程》
CAS
CSCD
北大核心
2007年第11期29-30,46,共3页
Computer Engineering
基金
上海市重点学科建设基金资助项目(T0602)
上海市教委科研基金资助项目
关键词
数据挖掘
频繁项集
双空间挖掘算法
关联规则
Data mining
Frequent items sets
Dual space mining algorithm
Associate rule
