EFFICIENTLY MINING HIGH AVERAGE-UTILITY ITEMSETS WITH AN IMPROVED UPPER-BOUND STRATEGY

摘要: Utility mining has recently been discussed in the field of data mining. A utility itemset considers both profits and quantities items transactions, thus its value increases with increasing length. To reveal a better effect, an average-utility measure, which is total divided by length, proposed. However, existing approaches use traditional upper-bound model to find high itemsets, generate large number unpromising candidates process. The present study proposes improved approach that uses prefix concept create tighter upper bounds values for reducing itemsets Results from experiments on two real databases show proposed algorithm outperforms other algorithms under various parameter settings.