The Community for Technology Leaders
Green Image
ISSN: 1041-4347
Kuldeep Singh , Computer science and engineering, Indian Institute of Technology (BHU) Varanasi, VARANASI, Uttar Pradesh India (e-mail: kuldeep.rs.cse13@iitbhu.ac.in)
Shashank Sheshar Singh , Department of Computer Science and Engineering, IIT (BHU), Varanasi, VARANASI, Uttar Pradesh India (e-mail: shashankss.rs.cse16@iitbhu.ac.in)
Ajay Kumar , Department of Computer Science and Engineering, IIT (BHU), Varanasi, VARANASI, Uttar Pradesh India (e-mail: ajayk.rs.cse16@iitbhu.ac.in)
Harish Kumar Shakya , Department of Computer Science and Engineering, IIT (BHU),Varanasi, VARANASI, Uttar Pradesh India (e-mail: hkshakya.rs.cse@iitbhu.ac.in)
Bhaskar Biswas , CSED, IIT-BHU, VARANASI, Uttar Pradesh India (e-mail: bhaskar.cse@iitbhu.ac.in)
ABSTRACT
High utility itemsets (HUIs) mining is a subfield of frequent itemsets mining. Traditional HUIs mining algorithms mine a large number of HUIs, but most of the mined HUIs are redundant. In order to overcome this issue, closed HUIs (CHUIs) mining algorithms have been proposed which avoids redundant itemsets. However, traditional CHUIs mining algorithms work only with positive utility. A CHUIs mining algorithm with negative utility has not yet been proposed, although negative utility mining is commonly seen in real-world applications. In order to address this issue, we propose an efficient algorithm named CHN (Closed High utility itemsets mining with Negative utility). It relies on a pattern-growth approach and utilizes transaction merging and dataset projection techniques to reduce the dataset scanning cost. In order to enhance the performance of CHN, we utilize sub-tree based pruning technique. A bi-directional extension technique is also utilized to check the closure and prune the search space. In order to check the efficiency of utilized techniques, the variation of the proposed algorithm named CHN (RSU-Prune) and CHN (TM) are introduced. The experimental results on the dense and sparse datasets show that the proposed algorithm and its variants outperform the state-of-the-art algolrithm.
INDEX TERMS
Itemsets, Data mining, Memory management, Merging, Redundancy, Data structures, Generators
CITATION

K. Singh, S. S. Singh, A. Kumar, H. K. Shakya and B. Biswas, "CHN: an efficient algorithm for mining closed high utility itemsets with negative utility," in IEEE Transactions on Knowledge & Data Engineering.
doi:10.1109/TKDE.2018.2882421
93 ms
(Ver 3.3 (11022016))