Reducing Database Scans for On-shelf Utility Mining

Utility mining has recently been an emerging topic in the field of data mining. It finds out high-utility itemsets by considering both the profits and quantities of items in transactions. In real applications, however, utility mining may have a bias if items are not always on shelf. On-shelf utility mining is then proposed, which considers not only individual profit and quantity of each item in a transaction but also common on-shelf time periods of a product combination. In the past, a two-phase on-shelf utility mining was proposed to discover the desired patterns in on-shelf utility mining. It, however, adopted the level-wise mining way to find the patterns. To speed up the execution efficiency, a three-scan mining approach is thus proposed in the paper to efficiently discover high on-shelf utility itemsets. The proposed approach utilizes an itemset-generation mechanism to prune redundant candidates early and to systematically check the itemsets from transactions. At last, the experimental results on synthetic datasets show the proposed approach has a better performance than the previous one.