Mining high utility patterns in different time periods

Abstract

Mining frequent itemset generates frequently purchased itemsets. However, a frequent itemset may not be the itemset with high value. Mining high utility itemset considers both of the profits and purchased quantities for the items, which is to find the itemsets with high utility for the business. However, mining high utility itemsets from the whole database may lose the high utility itemsets in a specific time period. Mining high utility itemsets in different time periods considers when the itemsets will be high utility, which can provide business manager to promote high utility itemsets in a specific time period. In this paper, we propose an approach for mining high utility itemsets in different time periods. We first partition the database according to the userspecified basic time periods, and then combine the transactions in the continuous basic periods to find high utility itemsets in the continuous periods. Besides, we remain the itemsets which are not high utility in the previous periods, but may be high utility after combining the previous periods with the current basic period, such that the high utility itemsets in continuous basic periods will not be lost.