摘要
关联规则挖掘技术能够从复杂数据中发现有意义的关联知识,但是目前还没有有效的执行算法,为了提高频集发现问题中的存在的执行效率问题,引入了基于关联图的数据表示技术,提出了基于关联图的频集快速发现算法(conjunctiongraph-basedfrequent-setsfastfindingalgorithm,CGFF),根据关联图的结构特性,有效地实现了频集发现的合理剪枝问题,大大提高了执行效率,最后通过实验证明频集快速发现算法是行之有效的。
There are no efficient algorithms to mining correlation rules in the field ofdata mining. To enhance the efficiency ofprocessing to find frequent sets, a novel algorithm for mining complete frequent itemsets is presented. This algorithm is referred to as the conjunction graph-based frequent-sets fast finding algorithm from hereon, In this algorithm, the graph-based pruning to produce frequent patterns is employed. Experimental data show that the CGFF algorithm outperforms than others.
出处
《计算机工程与设计》
CSCD
北大核心
2006年第17期3136-3139,共4页
Computer Engineering and Design
基金
广东省科技基金项目(2005B10101033
2004A10202001)。
关键词
数据挖掘
关联规则
频集
关联图
关联规则挖掘
data mining
correlation rules
frequent itemsets
conjunction graph
correlation rules mining