Extraction of high utility rare itemsets from transactional databases

Association rule mining is the task of data mining, which generates rules based on the relationships between the set of items purchased.We propose an algorithm, namely utility pattern rare itemset (UPRI), for mining high utility rare itemsets with a set of strategies. The information of high utility itemsets is maintained in a tree-based data structure namely utility pattern rare tree (UPR-Tree). Utility mining aims to discover itemsets with high utilities by considering profit, quantity, cost or other user preferences. In retail business high consideration should be given to utility of item in a transaction, since items having low selling frequencies may have high profits. Rare itemsets provide useful information in different decision making domains. In this paper, UPRI algorithm has been proposed to generate high utility rare itemsets. These itemsets occur infrequently in a transactional database but may generate huge profits for a business.