摘要： 在传统的多层关联挖掘算法中，概念层次树只提供项目的概念分层信息，导致项目查找效率不高。为此，提出一种改进的多层关联规则挖掘算法。在概念层次树的节点中增加2个域，以获取各层的频繁1-项集，通过增加1个基于Hash表的辅助存储结构，以提高项目的查找效率。实验结果表明，与传统算法相比，改进算法的挖掘时间能缩短10%左右。

关键词: 多层关联规则, 数据挖掘, 概念层次树, 频繁模式树

Abstract: In the traditional association rule mining algorithm, concept hierarchy tree only provides concept hierarchy information of items, and the efficency of searching items is not high in this tree structure. This paper presents an improved multi-level association rule mining algorithm. It adds two fields in its nodes to help to accquire frequent 1-itemsets. It employs an auxiliary storage structure that is based on Hash table, which enhances efficency of searching items. Experimental result shows that mining time of the imrpoved algorithm is reduced by about 10% than before.

Key words: multiple-level association rule, data mining, concept hierarchy tree, Frequent Pattern(FP) tree

中图分类号: 

• TP311.13