dMITP-Miner: An Efficient Method for Mining Maximal Inter-transaction Patterns

Abstract

In this paper, we propose an efficient algorithm, namely dMITP-Miner, for mining frequent maximal inter-transaction patterns (FMITPs). The proposed method uses diffset to store the information of patterns for efficiently mining FMITPs. In addition, we also proposed effective pruning strategies to help in reducing the search space to speed up the runtime and to cut down the memory usage. Experiments have been conducted to compare the effectiveness between the dMITP-Miner and the tMITP-Miner method in terms of runtime and memory usage.