Abstract: Mining high utility item sets from transactional databases refers to finding the item sets with high profits. Here, the meaning of item sets utility is interestingness, importance, or profitability of an item to users. A large number of candidate item sets for high utility item sets degrades the mining performance in terms of execution time and space requirement. In this thesis, an algorithm, namely utility pattern growth (UP-Growth) is used for mining high utility item sets with a set of effective strategies for pruning candidate item sets. The information of high utility item sets is maintained in a tree-based data structure which is named as utility pattern tree (UP-Tree) such that candidate item sets can be generated efficiently with scanning the database. To facilitate the mining performance and avoid scanning original database repeatedly, a compact tree structure, named UP-Tree is used, to maintain the information of transactions and high utility item sets. EEGU (Enhanced Eliminating Global Unpromising items) and EEGNU (Enhanced Eliminating Global Node Utilities) strategies are applied to minimize the overestimated utilities stored in the nodes of global UP-Tree. By applying strategy EEGU (Enhanced Eliminating Global Unpromising items), the utilities of the nodes that are closer to the root of a global UP-Tree are further improved. Experimental results show that the proposed algorithms, especially UP-Tree with EEGU and EEGNU, reduce time complexity rate. It also outperform in the dynamic transaction dataset.
Keywords: frequent item set, high utility item set, utility mining, data mining.
Title: High Utility Item sets Mining from Transactional Databases
Author: B.JAYANTHI, S.SARASWATHI
International Journal of Computer Science and Information Technology Research
ISSN 2348-1196 (print), ISSN 2348-120X (online)
Research Publish Journals
