一种不产生候选项集的关联规则挖掘算法

One kind does not have the candidate item of Association rule mining algorithm

在对Apriori算法的特性进行详细的介绍和总结的基础上,提出一种新的不产生候选项集的,以及只进行少量数据库扫描的挖掘频繁项集的算法SI-tree(Supper-Item Tree),该算法通过搜索数据库和一次性地找出当前项的所有超集来查找频繁项集.最后通过实验对2种算法进行了比较.

Based on detailed introduction to and on the summary of Apriori algorithm, this article proposes one new kind of SI_tree（Supper-Item Tree） that does not produce the candidate item of collection and just needs a few scanning and mining database. This algorithm, through the search database, discovers all ultra collections search at a time, and finally makes a comparison of the 2 Apriori through the experiment.