A Decremental Utility Mining Algorithm Based on the Pre-large Concept

抄録

In the past, utility mining was proposed to measure the utility values of purchased items for revealing high utility itemsets from a quantitative database. In dynamic data mining, transactions may be inserted, or deleted in the database. A batch mining procedure must rescan the whole updated database to maintain the up-to-date information. In this paper, a decremental mining algorithm is thus proposed for efficiently maintaining the discovered high utility itemsets due to transaction deletion based on the pre-large concept. Experimental results show that the proposed decremental high utility mining algorithm outperforms existing batch algorithms.