Maintenance of high average-utility itemsets for record deletion

Traditionally, the utility of an itemset is the summation of the utilities of the itemset in transactions regardless of its length. The average utility measure was then proposed to reveal a better utility effect of combining several items than the utility measure. In this paper, a maintenance algorithm for high average-utility itemsets is proposed for reducing execution time when records are deleted. The proposed approach is designed based on the two-phase utility mining and the FUP (Fast UPdate) concept and can achieve a good performance especially when the knowledge changes only to a small extent.