树形频繁模式的挖掘及可视化技术研究

Research on techniques of mining and visualizing frequent patterns based on tree structure

频繁项集的挖掘不仅仅是关联规则挖掘的基础,而且在序列模式、聚类、多维模式等数据挖掘任务中扮演重要角色.本文在给出一个基于数据垂直分布的频繁项集挖掘算法HBMFP的基础上,论述了利用MFC中的树视图控件(CTreeCtrl)将频繁项集树形可视化,并讨论了基于该频繁模式树的3种约束频繁项集查询的方法.

Mining of frequent itemsets is not only a fundamental problem for mining association rules,but also plays an improtant role in other data mining rasks such as sequential patterns,clusters,multidimensional patterns,etc.Based on the analysis of the vertical format-based frequent pattern mining algorithm HBMFP,in particular this paper visualizes the frequent patterns making use of CtreeCtrl in MFC.Discussing the three methods of querying of constraind frequent itemsets based on the frequent pattern tree.