计算机工程与科学2009，Vol.31Issue(12)：71-73,133,4.DOI:10.3969/j.issn.1007-130X.2009.12.021
一种基于压缩前缀树的频繁模式挖掘算法
A Mining Algorithm for Frequent Patterns Based on Compressed Prefix-Tree
郭云峰 1张集祥1
作者信息
- 1. 杭州电子科技大学图形图像研究所,浙江,杭州,310018
- 折叠
摘要
Abstract
The FP-growth algorithm achieves better performance and efficiency than the Apriori-like algorithms because of avoiding costly candidate generation, but it still suffers from creating conditional FP-trees separately and recursively during the mining process, so its efficiency in time and space is not idear. In this paper, we propose a new algorithm CPM that designs a new tree structure called compressed prefix-tree, which stores all of the information in a highly compact form, decreases the consumption of system resources greatly. CPM mines frequent patterns in a depth-first order and directly in compressed prefix-trees by adjusting the node information and node links without using any additional data structures. Thus, it improves performance greatly.关键词
频繁模式/压缩前缀树/频繁项集Key words
frequent pattern/compressed prefix-tree/frequent itermset分类
信息技术与安全科学引用本文复制引用
郭云峰,张集祥..一种基于压缩前缀树的频繁模式挖掘算法[J].计算机工程与科学,2009,31(12):71-73,133,4.
