摘要
发现频繁项集是关联规则挖掘中最基本、最重要的问题。目前已有两类频繁项集挖掘算法,然而由于其内在的复杂性,这一问题并未完全解决。提出了一种基于FP-Tree的频繁项集挖掘算法,该算法通过计算FP-Tree中非叶子节点的频繁子孙集和频繁前缀,组合生成频繁项集,无需递归构造每个频繁项的条件模式树,节约了时间和内存空间,算法性能在一定程度上得到了提高。
Discovery of frequent itemsets is the most fundamental and important issue in mining association rules.Although there are two types of mining algorithms to solve this problem currently,it has not been fully addressed yet due to the inherent computational complexity.In this paper,we present an algorithm of frequent itemsets mining based on FP-Tree.The algorithm combinatorially generates frequent itemsets through calculating the frequent-posterity and frequent-prefix of non-leaf nodes of the FP-Tree.The recursive construction of conditional model FP-Tree of each frequent item is not necessary,which saves the time consumed and the memory space,so the performance of the algorithm can be improved to some degree.
出处
《计算机应用与软件》
CSCD
2010年第10期36-37,130,共3页
Computer Applications and Software
基金
国家自然科学基金项目(60873234)
