Algorithm for generating strong association rules based on matrix
一种基于矩阵的强关联规则生成算法*

Apriori, the classic association rule mining algorithm, has the problems of higher I/O cost and more candidate itemsets in the process of finding out the frequent itemsets. Here, a new algorithm generating strong association rules based on matrix is proposed. The new algorithm could only scan the database for one time to convert the transactions into matrix which is composed by 0 and 1, and can be reordered by item support count non-descending order to reduce the number of candidate itemsets, meanwhile, the efficiency of association rule confidence computing can be enhanced too. The analysis results of examples and large database show that the proposed method is effective.