Method for Mining Frequent Item Sets Considering Average Utility
Reshu Agarwal, Arti Gautam, Ayush Kumar Saksena, Amrita Rai, Shylaja Vinaykumar Karatangi
Abstract
Data mining is used to determine helpful data and information from large data bases for decision making in different fields. In today's scenario, mining high-utility itemsets (HUIs) has become a current investigation area. In traditional method HUIs are found on the basis individual utility of an item-set, which in turn is calculated as the totality of the utilities of individual items. But the problem is that the above way of finding HUIs does not consider length of item-set into consideration, while calculating HUIs. In real world situations, it is desirable to calculate average utility while considering both the length of item-sets and their utilities. This paper proposes an algorithm to find the high average-utility item-sets (HAUIs) seeing both parameters i.e. length of itemsets and their utilities. A numerical example is devised to explain the proposed approach. Experiments on real life databases shows that suggested model can proficiently discover the total set of HAUIs.