分析了Apriori算法关于发现频繁项集的方法及其效率,提出了一种基于上三角项集矩阵的频繁项集挖掘优化算法。本算法只需要扫描数据库一次,不产生候选项目集,也不使用逐层迭代的方法,大大提高了频繁项集的发现效率。
This article proposes an algorithm of frequent itemsets mining based on Upper Triangular Itemsets Matrix ( UTIM ), by analyzing the way and efficiency that Apriori algorithm discovers frequent itemsets. The algorithm scans database only once, does not create candidate itemsets, and does not use the method of iteration for each layer. The efficiency is distinctly improved in discovers frequent itemsets.