Efficient Method for Mining High-Utility Itemsets Using High-Average Utility Measure.

Mining high-utility itemsets (HUIs) based on high-average utility measure is an important task in the data mining field. However, many of the existing algorithms are performing the mining process sequentially and do not utilize the widely available multi-core processors, thus requiring long execution times. To address this issue, we propose an extended version of the HAUI-Miner algorithm, namely pHAUI-Miner. The algorithm applies multi-thread parallel processing to significantly reduce the mining time. Experimental evaluations on standard databases have shown the effectiveness of the proposed algorithm over the original and sequential method.