基于频繁模式树的快速关联规则挖掘算法

摘要

提出了一种挖掘频繁项目集的有效算法--FFP-Growth,该算法采用自底向上的策略搜索频繁模式树,但不同于FP-Growth的是它无须生成条件模式基和频繁模式子树,且生成的频繁模式树较TD-FP-Growth生成的频繁模式树小,因而能提高关联规则的挖掘效率. 类似于TD-FP-Growth的扩展TD-FP-Growth(M) 和TD-FP-Growth(C),FFP-Growth很容易被扩展,以此来有效地减小搜索空间. 实验结果表明本文提出的算法是有效可行的.%In this paper, we propose an efficient algorithm, called FFP-Growth (short for fast FP-Growth), to mine frequent itemsets. Similar to FP-Growth, FFP-Growth searches the FP-tree in the bottom-up order, but need not construct conditional pattern bases and sub-FP-trees, thus, saving a substantial amount of time and space, and the FP-tree created by it is much smaller than that created by TD-FP-Growth, hence improving efficiency. At the same time, FFP-Growth can be easily extended for reducing the search space as TD-FP-Growth (M) and TD-FP-Growth (C). Experimental results show that the algorithm of this paper is effective and efficient.