ITWUI - An Interactive Transaction Weighted Utility Item Tree for High Utility Itemset Mining

Abstract

Currently in the competitive marketing era, high utility itemset (HUI) mining is the most useful and most advanced research area which inherits from association rule mining (ARM). HUI is useful for designing the strategies for the big retailer mall, e-commerce portal like amazon, alibaba, ebay, flipkart, etc.. Utility in a high utility itemset is a profit which is calculated from the product of quantity and price of each product from the transaction table. In this paper, we propose a  matrix called transaction positioning matrix. Transaction positioning matrix is a tabular data structure which is derived from a transaction database.   Based on this matrix, we have generated a tree for each item from the transaction named as interactive transaction weighted utility item (ITWUI) tree.  ITWUI tree is created for each potential item present in the transaction database. Then finally, a high utility itemset is mined by processing the ITWUI tree in parallel.