基于频繁模式树的约束最大频繁项集挖掘算法  

Mining Algorithm for Constrained Maximum Frequent Itemsets Based on Frequent Pattern Tree

作　　者：花红娟[1] 张健[2] 陈少华[1] 

机构地区：[1]上海海洋大学信息学院,上海201306 [2]上海海洋大学图书馆,上海201306

出　　处：《计算机工程》2011年第9期78-80,共3页Computer Engineering

基　　金：国家"863"计划基金资助重点项目"人工鱼礁生态增殖及海域生态调控技术"(2006AA100303)

摘　　要：多数最大频繁项集挖掘算法产生候选项目集的代价很高,而实际应用中用户只关心部分关联规则。针对该问题,提出一种基于频繁模式树的约束最大频繁项集快速挖掘算法。该算法能随时删除不满足约束条件的项集,无需生成候选项目集,由此提高挖掘效率。实验结果证明,该算法的效率优于同类算法。The cost of producing candidate itemsets is very high in most maximum frequent itemset mining algorithms,but users are often interested in a subset of association rules in practical application,so this paper proposes a mining algorithm for constrained maximum frequent itemsets based on Frequent Pattern tree（FP-tree）.It can delete the itemsets which do not meet the constraints at any time and does not produce candidate itemsets,so that the efficiency of mining is improved.Experimental results show that the algorithm is better than other algorithm.

关 键 词：数据挖掘 最大频繁项集 约束最大频繁项集 频繁模式树 项约束 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]