A NOVEL APPROACH FOR MINING INTER-TRANSACTION ITEMSETS

Abstract

Aim: To address the problem of inter-transaction association rule mining, where the frequent occurrence of a large number of items results in a combinatorial explosion that limits the practical application of the existing mining algorithms. Methodology: We propose an efficient algorithm called IAR Miner (Inter-transaction Association Rule Miner), for mining inter-transaction itemsets. Our proposed algorithm consists of two phases. First, we scan the database once to ※nd the frequent items. For each frequent item found, the IAR Miner converts the original transaction database into a set of domain attributes, called a dataset. Then, it enumerates inter-transaction itemsets using an Itemset-Dataset tree, called an ID-tree. By using the ID-tree and datasets to mine inter-transaction itemsets, the IAR Miner can embed effective pruning strategies to avoid costly candidate generation and repeated support counting. Results: Our proposed algorithm can efficiently mine inter-transaction patterns. The performance study on the synthetic datasets shows that the IAR Miner algorithm is more efficient than the EH-Apriori, FITI, ClosedPROWL and ITP-Miner algorithms in most cases. Conclusion: The IAR Miner algorithm can efficiently mine the inter-transaction patterns. In the future work, we will address a number of research issues related to the IAR Miner algorithm.