Cải tiến thuật toán Hminer cho việc khai thác tập hữu ích cao trên dữ liệu thao tác thưa

Abstract

High-utility itemset mining plays an important role in data mining. This mining helps to discover highly useful itemsets, i.e., itemsets of high importance or profit, in transactional databases. That helps companies and supermarkets to give appropriate business orientation and strategies to bring the highest profit. Depending on the form of dense or sparse data, mining algorithms will have suitable mining strategies and achieve certain efficiency. This paper focuses on researching and proposing mining methods on sparse datasets through data presentations and pruning techniques. The experimental evaluation results have proved the feasibility of the proposed solution.