Computer Integrated Manufacturing System ›› 2024, Vol. 30 ›› Issue (8): 2631-2643.DOI: 10.13196/j.cims.2023.BPM01

    Next Articles

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

FU Jianping1,ZHAO Haiyan1+,CAO Jian2,CHEN Qingkui1   

  1. 1.School of Optoelectronic Information and Computer Engineering,University of Shanghai for Science and Technology
    2.School of Electronic Information and Electrical Engineering,Shanghai Jiao Tong University
  • Online:2024-08-31 Published:2024-09-03
  • Supported by:
    Project supported by the National Natural Science Foundation,China(No.62072301).

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

付建平1,赵海燕1+,曹健2,陈庆奎1   

  1. 1.上海理工大学光电信息与计算机工程学院
    2.上海交通大学电子信息与电气工程学院
  • 作者简介:
    付建平(1995-),男,江西赣州人,硕士研究生,研究方向:深度学习、数据挖掘,E-mail:jianpingfu@163.com;

    +赵海燕(1975-),女,河南焦作人,副教授,博士,硕士生导师,研究方向:服务计算、数据挖掘、推荐系统,通讯作者,E-mail:zhaohaiyan1992@foxmail.com;

    曹健(1972-),男,江苏宜兴人,教授,博士,博士生导师,研究方向:智能数据分析、协同计算、服务计算、网络计算等,E-mail:cao-jian@sjtu.edu.cn;

    陈庆奎(1967-),男,黑龙江哈尔滨人,教授,博士,博士生导师,研究方向:计算机集群、并行数据库、并行理论、物联网等,E-mail:chenqingk@tom.com。
  • 基金资助:
    国家自然科学基金资助项目(62072301)。

Abstract: 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.

Key words: business process, event log, control flow, anomaly detection

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。

关键词: 业务过程, 事件日志, 控制流, 异常检测

CLC Number: