The Research on Association Rules Algorithm Based on Minimum Item Supports
Xiao-Gao Yu · 2008
Association rules algorithm is very important in data mining. Apriori algorithm is analyzed, which is classic one in the association rules algorithms and summarizes problems existing in the algorithm. Study the frequent itemsets problem for association rules in data mining and a new association rules algorithm based on minimum item supports called MSOA is proposed. In this algorithm, the itemsets are ordered by ascending order instead by lexicographic order. It can greatly reduce the candidate frequent itemsets, keeps the completion of frequent itemsets, and reduces the cost of computing. Experiment results show that the algorithm is a high efficient algorithm which can mine all the frequent itemsets by scanning the source database only once.