An Improved Approach For High Utility Itemset Mining Using Length Reduction Method

Abstract

Data mining is process of analyzing data from different data repository and mine Useful and interesting patterns from them. It is also defined as the use of algorithm to discover hidden patterns and interesting relationship between large itemset. High utility itemset mining is an area research where utility based mining can be done. Mining high utility itemset from a transactional database refers to the discovery of itemset with high utility in a terms like weight, unit profit or value. High-utility item set mining (HUIM) is an important data mining task that refers to the set of items with high utility like profit in a customer transaction database. However an Important issue with traditional HUI mining algorithm is that they tend to find itemset having many items which increases memory and time overhead. To discover HUIs efficiently with length constraints, FHM+ introduced the concept of estimated utility co-occurrence structure (EUCS) and two Length Upper Bound Reduction (LUR) of itemset. EUCS has matrix structure and in that half of the matrix is not filled with data so it has memory overhead.. In this paper, we address this issue by presenting an improved algorithm based on tree data structure which can decreases the execution time and memory usage for HUI mining.