International Journal of Science and Research (IJSR)

International Journal of Science and Research (IJSR)
Call for Papers | Fully Refereed | Open Access | Double Blind Peer Reviewed

ISSN: 2319-7064


Downloads: 109

India | Computer Science Engineering | Volume 5 Issue 11, November 2016 | Pages: 1419 - 1424


High Utility Itemsets Mining ? A Brief Explanation with a Proposal

Anu Augustin, Dr. Vince Paul

Abstract: High utility itemsets mining is relevant for business vendors. So that they can give more offers to high utility itemsets. To understand the above sentence we need to know what is high utility itemsets. High utility itemsets are those ones that yield high profit when sold together or alone that meets a user-specified minimum utility threshold from a transactional database. This high utility itemset mining is not a new topic, but it is an emerging area. The basis of high utility mining is frequent itemset mining. The various problems in frequent itemset mining are purchase quantity not taken into account, all items have same importance etc. So the number of items generated will be more. These limitations are overcomed by high utility itemset mining. For that in HUI mining a utility value (weight) is assigned to each item. Also a threshold applied to remove unwanted itemsets. Setting the threshold externally is a tedious work. Too low threshold will generate many HUIs and too high may cause no HUIs to found. In Top-K only top hui s will be found. Here the minimum threshold is set internally. It is zero initially. Performance degrades when there are many huis in the database. So the concept of closed itemset mining is introduced for memory and space efficiency. Also for proper utilization of resources.

Keywords: Frequent Pattern, High Utility Itemsets, Transaction Utility, Minimum Utility Threshold, Closed itemsets



Citation copied to Clipboard!

Rate this Article

5

Characters: 0

Received Comments

No approved comments available.

Rating submitted successfully!


Top