一种可交叉的带时间约束的互异最小情节计数方法  

A Cross-over Method for Counting Mutually Distinct Minimum Episodes with Time Constraints

作　　者：李威[1] 吴志刚 李健俊[1] 陆海龙[1] Li Wei;Wu Zhigang;Li Jianjun;Lu Hailong(China Tobacco Zhejiang Industrial Co.,Ltd.,Hangzhou 310001;Hangzhou Shiping Information&Technology Co.,Ltd.,Hangzhou 310012)

机构地区：[1]浙江中烟工业有限责任公司,杭州310001 [2]杭州世平信息科技有限公司,杭州310012

出　　处：《信息安全研究》2022年第12期1187-1191,共5页Journal of Information Security Research

摘　　要：情节挖掘中通常存在长串序列数据不易进行分割的现象,针对现有情节挖掘算法中未考虑时间约束、情节交叉发生以及存在重复计数等问题,提出一种可交叉的带时间约束的互异最小情节计数方法.综合考虑情节发生的时间跨度问题、信号事件的互异性以及情节的可交叉性,通过ONCE-TDM情节计数算法实现了单遍扫描序列数据,完成情节发生的精准计数.In the episode mining, it is usually difficult to segment long sequence data.Aiming at the problems that time constraints, cross occurrence of episodes, and repeated counting are not considered in the existing episode mining algorithms. We propose a cross-over method for counting mutually distinct minimum episodes with time constraints which considers the time span of the occurrence of the episode, the heterogeneity of signal events and the cross ability of the episode. By using the ONCE-TDM episode counting algorithm, we can scan the sequence data in a single pass and complete the accurate counting of the occurrence of the episode.

关 键 词：情节挖掘 时间约束 最小发生 序列数据 情节计数 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]