Implementation of Efficient Algorithms for Mining Top-K High Utility Item sets

Popular problem in data mining, which is called 'high-utility itemset mining' or more generally utility mining. High Utility Itemsets which are itemsets having a utility meeting a user-specified minimum utility threshold value i.e min_util. The main objective of utility mining is to find item sets with highest utilities , by considering profit, quantity, cost or any other user preferences. Research has been carried out in area of mining HUI's. Various techniques have been applied. The main problem with setting threshold value which is mostly user specific, is it needs to be appropriate. In Order to set most appropriate or right Threshold value for mining HUI's ,user needs to do trial & error which in turn is time consuming & tedious process, because if min_util is set too low , system will result in getting large data of HUI , which in turn makes system ineffective for the purpose of HUI. If we set min_util too high , this will result in getting small amount or no HUI's. Thus setting minimum threshold value is difficult. The proposed system is following Top-k framework for mining top-k HUI's, which is using two algorithms TKU (mining top-k utility itemsets) & TKO (mining top-k in one phase),without setting min_util threshold.