燕山大学学报2025，Vol.49Issue(4)：349-366,18.DOI:10.3969/j.issn.1007-791X.2025.04.008
一种快速的高效用序列模式挖掘算法
A fast algorithm for high utility sequential pattern mining
摘要
Abstract
The primary objective of high utility sequential pattern mining tasks is to extract high utility subsequences from sequence databases to acquire potential knowledge.However,the computation of utility for sequence data coupled with the explosion of search space generated by low utility thresholds makes high utility sequential pattern mining highly challenging.To address the issue of excessive time and memory consumption in existing high utility sequential pattern mining algorithms,a compact utility index list structure is proposed to store information such as the utility and position of sequences generated during the mining process.Based on this structure,a fast high utility sequential pattern mining algorithm is designed.To further enhance the mining efficiency of the algorithm,a new upper bound is proposed to reduce the search space.Extensive experiments on real and synthetic datasets demonstrate that the proposed algorithm outperforms state-of-the-art algorithms in terms of time,memory usage,search space reduction,and scalability.关键词
模式挖掘/高效用序列模式/序列分析/效用挖掘/效用索引列表Key words
pattern mining/high utility sequential patterns/sequence analysis/utility-oriented mining/utility index list分类
信息技术与安全科学引用本文复制引用
张瑞华,韩萌,何菲菲,孟凡兴,李春鹏..一种快速的高效用序列模式挖掘算法[J].燕山大学学报,2025,49(4):349-366,18.基金项目
国家自然科学基金资助项目(62062004) （62062004）
宁夏自然科学基金资助项目(2023AAC03315) （2023AAC03315）
北方民族大学中央高校基本科研业务费专项资金资助项目(2021KJCX10) （2021KJCX10）
北方民族大学研究生创新项目(YCX24120) （YCX24120）
