Research on Construction Method of Association Rule Mining Based on the Hash-tree

摘要

Association rule is one of the fundamental methods of data mining, the hash-tree is an important data structure that used to mining frequent itemset by homo-Apriori algorithms. Because the hash-tree can't be built successfully one-time, this paper proposes static and dynamic two hash-tree construction methods. Both methods can not only determine the hash table size, hash functions and the number of itemset in the leaf nodes, but also guarantee to build a hash-tree successfully one-time in any case and prove the correctness of the proposed method.