HDSHUI-miner: a novel algorithm for discovering spatial high-utility itemsets in high-dimensional spatiotemporal databases
HDSHUI-miner: a novel algorithm for discovering spatial high-utility itemsets in high-dimensional spatiotemporal databases
复制标题
HDSHUI-miner：一种在高维时空数据库中发现空间高效用项集的新算法
DOI：
10.1007/s10489-022-04436-w
复制
发表时间：
2023
影响因子：
5.3
通讯作者：
Bommisetti Sai Chithra
中科院分区：
文献类型：
--
作者：
Uday Kiran Rage;Veena Pamalla;Ravikumar Penugonda;Venus Vikranth Raj Bathala;Dao Minh-Son;Zettsu Koji;Bommisetti Sai Chithra
空间高效用项集挖掘是一种重要的大数据分析技术。它的目标是在时空数据库中定位所有具有高实用价值的地理上感兴趣的项集。文献中提出了一种求取期望项集的Shui-Miner算法。遗憾的是，该算法在处理高维时空数据库时存在性能问题。基于这一发现，本文扩展了最新的方法，提出了一种新的算法，称为高维水挖掘算法(HDSHUI-Miner)。我们的算法探索了几种新的剪枝策略，以减少寻找期望项目集所需的搜索空间和计算成本。在7个真实数据库上的实验结果表明，HDSHUI-Miner在内存消耗、运行时间和可伸缩性方面都优于Shui-Miner。最后，我们给出了两个真实世界的案例研究来说明该算法的有效性。
Spatial high-utility itemset (SHUI) mining is a significant big data analysis technique. It aims to locate all geographically interesting itemsets with high utility in a spatiotemporal database. An SHUI-Miner algorithm was presented in the literature to find the desired itemsets. Unfortunately, this algorithm suffered from performance issues when dealing with high-dimensional spatiotemporal databases. Based on this finding, this paper extends the state-of-the-art method by proposing a novel algorithm known as the high-dimensional SHUI-miner (HDSHUI-Miner). Our algorithm explores several novel pruning strategies to decrease the search space and computational cost required to find the desired itemsets. Experimental results obtained on seven real-world databases demonstrate that HDSHUI-Miner outperforms SHUI-Miner with respect to memory consumption, runtime, and scalability. Finally, we present two real-world case studies to illustrate the usefulness of the proposed algorithm.