基于事件日志的业务过程控制流异常检测算法:现状与评测  

Anomaly detection algorithm for business process control flow based on event log:Status and evaluation

作　　者：付建平 赵海燕[1] 曹健[2] 陈庆奎[1] FU Jianping;ZHAO Haiyan+;CAO Jian;CHEN Qingkui(School of Optoelectronic Information and Computer Engineering,University of Shanghai for Science and Technology,Shanghai 200093,China;School of Electronic Information and Electrical Engineering,Shanghai Jiao Tong University,Shanghai 200030,China)

机构地区：[1]上海理工大学光电信息与计算机工程学院,上海200093 [2]上海交通大学电子信息与电气工程学院,上海200030

出　　处：《计算机集成制造系统》2024年第8期2631-2643,共13页Computer Integrated Manufacturing Systems

基　　金：国家自然科学基金资助项目(62072301)。

摘　　要：在业务过程的执行过程中,各个视角都可能出现异常,如控制流异常、时间异常、数据流异常等。控制流是业务过程的主干,控制流的异常通常也隐含其他视角发生了异常,因此控制流异常的检测对于业务过程的正确执行具有关键性作用。近年来,越来越多的研究基于业务过程的事件日志检测控制流异常,针对这些检测方法,进行了分类和总结。此外,对典型的控制流检测方法进行实验,并分析实验结果,总结各种检测方法的优缺点及其适用场景,以帮助业务过程分析人员根据特定的场景选择合适的检测方法。During the execution of the business process,it is prone to various anomalies,such as control flow anomaly,time anomaly,and data flow anomaly.Control flow is the backbone of a business process,and anomalies in control flow usually imply anomalies from other perspectives.Therefore,the detection of control flow anomaly plays a key role in the correct execution of business processes.In recent years,more and more researches had been carried out on the detection of control flow anomaly based on business process event logs,and these detection methods were classified and summarized.In addition,experiments and results analysis of typical control flow detection methods were carried out,and the advantages and disadvantages of various detection methods and their applicable scenarios were summarized,so that business process analysts could choose appropriate detection methods according to specific scenarios.

关 键 词：业务过程 事件日志 控制流 异常检测 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]