Mining Algorithm of Maximal Frequent Itemsets Based on Position Lattice

Maximal frequent itemsets mining is a fundamental and important problem in many data mining applications. In this paper, we present GMPV, a depth first search algorithm, which accurately displays itemset based on position vector, for mining maximal frequent itemsets. In GMPV algorithm, the transaction database is mapped to a Boolean matrix. The methods superset checking and pruning based on support are also used to increase the algorithm efficiency. Our experiment results show that GMPV algorithm is very validity.