MINING TOP-K HIGH UTILITY ITEMSETS AND FREQUENT PATTERNS IN ONE PHASE

Computer Science And Engineering

NATIONAL CONFERENCE ON TECHNOLOGICAL ADVANCEMENTS IN ENGINEERING

High utility itemsets (HUIs) mining is an trending topic in data mining,which refers to discovering all itemsets having a utility meeting a user-specified minimum utility threshold min_util.HUI aimed at finding itemsets that contribute most to the total utility.However, setting min_util appropriately is a difficult problem for users. Although a number of relevant algorithms have been proposed in recent years,but the problem is they produce large number of candidate itemsets for high utility itemsets.Generally utility mining adopt a two-phase, candidate generation approach,in the first phase,candidates of high utility patterns in the found out and in second phase, scan raw data one more time to identify high utility patterns from the candidates. The problem is that the number of candidates can be huge, which is the scalability and efficiency bottleneck.To solve this problem ,finds high utility patterns in a single phase without generating candidates.In this paper,address the above issues by proposing a new framework for top-k high utility itemset mining.TKO (mining Top-K utility itemsets in One phase) are proposed for mining itemsets without setting the min_util. Along with this frequent pattern is also generated using CR-Tree. CR Tree data structure is generated and then by using FP growth algorithm top-k frequent patterns is mined.Then the customer segmentation is done.Cluster the product based on customer.Then CBA (Customer Behaviour Analysis)is done and also emerging trends is given to the customers.By enabling companies to target specic groups of customers,find What are the top-k sets of products (i.e., itemsets) that contribute the highest profits to the company and according to the frequent pattern mining result, the merchandise in the supermarket is arranged together in the same place well-suited for customer.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.