Mining Algorithm for Constrained Maximum Frequent Itemsets Based on Frequent Pattern Tree

The cost of producing candidate itemsets is very high in most maximum frequent itemset mining algorithms,but users are often interested in a subset of association rules in practical application,so this paper proposes a mining algorithm for constrained maximum frequent itemsets based on Frequent Pattern tree(FP-tree).It can delete the itemsets which do not meet the constraints at any time and does not produce candidate itemsets,so that the efficiency of mining is improved.Experimental results show that the algorithm is better than other algorithm.