Mining of High Utility Item sets from Transactional Databases
D. Usha Nandini1, Ezil Sam Leni2, M. Maria Nimmy3
1D. Usha Nandini, Asst. Prof & Research Scholar, Department of CSE, Sathyabama University, Chennai, (Tamil Nadu), India.
2S. Dr. Ezil Sam Leni, Asst. Prof & HOD, Department of CSE, SRR Engineering college, Chennai, (Tamil Nadu), India.
3T. M. Maria Nimmy, Pursuing M.E in CSE at Sathyabama University, Chennai, (Tamil Nadu), India.
Manuscript received on March 21, 2014. | Revised Manuscript received on April 06, 2014. | Manuscript published on April 30, 2014. | PP: 17-21 | Volume-3, Issue-4, April 2014. | Retrieval Number: C2717023314/2013©BEIESP
Open Access | Ethics and Policies | Cite
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Efficient discovery of high utility item sets from transactional databases crucial task in data mining. UP-Growth and UP-Growth+ algorithms are proposed for mining high utility item sets. In this paper we also proposed a compact tree structure, called Utility pattern tree (UP-Tree) and it maintains the information of high utility item sets. Previously we proposed FP-Growth algorithm for mining only large number of frequent item sets, but not generate the high utility item sets. They have the issue of producing large number of candidate item sets and probably it degrades mining performance in terms of speed and space requirement. However, our previous study needs more space and execution time. Many algorithms are used to show the performance of UP-Growth and UP-Growth+. UP-Growth and UP-Growth+ becomes more efficient since database contain long transactions and generate fewer number of candidates than FP-Growth. The experimental results and comparison validate its effectiveness.
Keywords: Candidate pruning, Data mining, Frequent itemset, High utility itemset.