摘要
针对传统数据关联挖掘过程只适用于单段数据集,导致内存负担重、挖掘频繁项集效率不高等问题,提出一种多段支持度数据频繁模式关联规则挖掘方法。运用多支持度算法对数据集逐步搜索,数据集按照数据项的MIS大小有序排列,采用最小值作为最小支持度,确保该算法的地推性。构建FPtree树,利用FPtree算法对待选项实施剪枝,从而准确挖掘出频繁模式的关联规则。仿真结果证明,多段支持度数据频繁模式关联规则挖掘具有较好的性能,有效提高了关联规则的挖掘效率。
In the traditional data association mining process,single segment data set leads to low efficiency of mining frequent itemsets.Therefore,a method of mining association rules in the frequent patterns of multi-segment support data was proposed.The multi-support algorithm was used to search the data sets step by step.According to the MIS of data items,data sets were arranged in order.The minimum value was adopted as the minimum support to ensure the recursion of the algorithm.Then,an FPtree tree was constructed,and the FPtree algorithm was used to prune,and thus to accurately mine the association rules of the frequent patterns.Simulation results show that the mining of association rules in the frequent patterns of multi-segment support data has a better performance,which improves the mining efficiency of association rules.
作者
王培培
孟芸
WANG Pei-pei;MENG Yun(Minsheng College,Henan University,Kaifeng Henan 475000,China)
出处
《计算机仿真》
北大核心
2021年第5期282-286,共5页
Computer Simulation
关键词
多段支持度
频繁模式
关联规则
数据挖掘
数据集缩减
Multi segment support
Frequent pattern
Association rule
Data mining
Data set reduction