An Innovative Procedure for Efficient Mining of Closed Top-K High Utility Itemsets
J. Wisely Joe1, S. P. Syed Ibrahim2
1J.Wisely Joe*,SCSE, VIT University (Chennai Campus), India.
2Dr.S.P.Syed Ibrahim, SCSE, VIT University (Chennai Campus), India.
Manuscript received on January 02, 2020. | Revised Manuscript received on January 15, 2020. | Manuscript published on January 30, 2020. | PP: 775-781 | Volume-8 Issue-5, January 2020. | Retrieval Number: E5737018520/2020©BEIESP | DOI: 10.35940/ijrte.E5737.018520
Open Access | Ethics and Policies | Cite | Mendeley
© 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: High utilization itemset (HUI) mining is the fastest growing ground in association finding between the items. It is a process of finding the itemsets with higher utility values which participates in profitable decision making. The generated HUIs reflect the frequency, importance, profit or the utility of the items present in the database. Proper minimum threshold setting is very difficult for the end users without the knowledge of the data present in the database. Minimum user threshold extracts more number of candidate sets . Higher user threshold gives less number of candidate sets and very few high utility itemsets . In both the cases, the process is inefficient. Some algorithms produce more number of candidate itemsets as HUIs. The set of HUIs may degrade the performance of the candidate set mining by increasing the storage and time when the database has very large number of transactions. The number of candidate itemsets involved in the generation of HUIs may also slow down the entire process. The proposed novel strategy for tapping top-k closed high utility itemsets out of the set of candidate sets addresses these issues . The user defined integer k is the needed count of HUIs to be extracted out of the quality itemsets. The algorithm does not require the user to set the minimum utilization threshold .The closure property is merged with the pruning process and improves the productivity. The results transparently show that the k-closed high utility itemsets generated using this algorithm are productive, profitable and very concise when compared with existing approaches.
Keywords: Association , Utility Mining, Utility Pattern Tree, High Utilization Itemset, Closed Property.
Scope of the Article: Text Mining.