Efficient Algorithms for Mining High Utility Itemset

Data mining uses various algorithms for searching interesting information and hidden patterns from the large database. Traditional frequent itemset mining (FIM) generate large amount of frequent itemset without considering the quantity and profit of item purchased. High utility itemset mining (HUIM) gives advantageous results as compared to the frequent itemset mining. HUIM algorithm helps to improve the performance of finding data by considering both quantity and profit of itemset from large database. This paper reviews two types of efficient algorithm named TKU (mining top-k utility itemset) and TKO (mining top-k utility itemsets in One phase) for mining high utility itemset without any need to set minimum utility threshold by using strategy of UP-tree data structure which scans the database twice and enhances the efficiency of mining High utility itemset. It find out transaction utility of each transaction and it also compute TWU of each item. Then it reorganizes the transaction and constructs the Up Tree.