Title: UP-GNIV: an expeditious high utility pattern mining algorithm for itemsets with negative utility values
Authors: Kannimuthu Subramanian; Premalatha Kandhasamy
Department of CSE, Coimbatore Institute of Engineering and Technology, Coimbatore-641109, Tamil Nadu, India
Department of CSE, Bannari Amman Institute of Technology, Sathyamangalam-638401, Tamil Nadu, India
Abstract: Traditionally, frequent pattern mining dealt in extracting frequency pattern from transaction databases by not considering utility factors. Utility-based data mining focuses on all aspects of economic utility in data mining and is aimed at incorporating utility in both predictive and descriptive data mining tasks. High utility itemset (HUI) mining process incurs the problem of producing a large number of candidate itemsets since downward closure property used in frequent itemset mining is not applied in utility mining and itemsets associated with negative utility values are not supported by existing algorithms. Here, mining high utility itemset with negative item values using Utility Pattern-Growth approach for Negative Item Values (UP-GNIV) approach is proposed and compared against high utility itemsets with negative item values (HUINIV)-mine. The experimental result shows that the suggested approach performs well.
Keywords: association rules mining; ARM; high utility itemsets; utility mining; negative item values; pattern mining; data mining.
Int. J. of Information Technology and Management, 2015 Vol.14, No.1, pp.26 - 42
Available online: 30 Nov 2014