Apriori核心算法过程如下：



过单趟扫描数据库D计算出各个1项集的支持度，得到频繁1项集的集合。

连接步：为了生成，预先生成,由2个只有一个项不同的属于的频集做一个（k-2）JOIN运算得到的。

剪枝步：由于是的超集，所以可能有些元素不是频繁的。在潜在k项集的某个子集不是中的成员是，则该潜在频繁项集不可能是频繁的可以从中移去。

通过单趟扫描数据库D，计算中各个项集的支持度，将中不满足支持度的项集去掉形成。-Apriori core algorithm process is as follows:





After single trip brain-scan database D calculated each 1 itemsets support and get frequent 1 itemsets collection.



Connect step: in order to create, a predetermined by 2 generate, only one item different belongs to the frequency of the set a (k- 2) JOIN operation get.



The pruning step: because yes superset and may, therefore, some elements are not frequent. In potential k itemsets is a subset of the members of the potential is, the frequent itemsets impossible is the frequent can find removed.



Through a single trip brain-scan database D, calculation of various itemsets support and will not satisfy the support degree itemsets removed form.