TDUP: an approach to incremental mining of frequent itemsets with three-way-decision pattern updating

Finding an efficient approach to incrementally update and maintain frequent itemsets is an important aspect of data mining. Earlier incremental algorithms focused on reducing the number of scans of the original database while it is updated. However, they still required the database to be rescanned in some situations. Here we propose a three-way decision update pattern approach (TDUP) along with a synchronization mechanism for this issue. With two support-based measures, all possible itemsets are divided into positive, boundary, and negative regions. TDUP efficiently updates frequent itemsets online, while the synchronization mechanism is periodically triggered to recompute the itemsets offline. The operation of the mechanism based on appropriate settings of two support-based measures is examined through experiments. Results from three real-world data sets show that the proposed approach is efficient and reliable.