计算机集成制造系统 ›› 2019, Vol. 25 ›› Issue (第4): 901-908.DOI: 10.13196/j.cims.2019.04.012

• 当期目次 • 上一篇    下一篇

基于后继关系的行为块过程挖掘方法

方欢,段瑞,詹悦   

  1. 安徽理工大学数学与大数据学院
  • 出版日期:2019-04-30 发布日期:2019-04-30
  • 基金资助:
    国家自然科学基金资助项目(61472003,61272153,61340003,61402011,61572035);安徽省自然科学基金资助项目(1608085QF149);安徽省高校优秀青年人才支持项目(gxyqZD2018038);安徽省博士后基金资助项目(2018B288)。

Process mining approach of discovering behavior blocks based on successor relation

  • Online:2019-04-30 Published:2019-04-30
  • Supported by:
    Project supported by the National Natural Science Foundation,China(No.61472003,61272153,61340003,61402011,61572035),the Natural Science Foundation of Anhui Province,China(No.1608085QF149),the Youth Talents Support Program in Universities of Anhui Province,China(No.gxyqZD2018038),and the Postdoctoral Funds of Anhui Province,China(No.2018B288).

摘要: 鉴于已有的过程挖掘方法在发现循环结构和隐藏行为上具有一定的局限性,提出一种基于后继关系的行为块挖掘算法。依据日志建立后继关系矩阵,分析了矩阵中变迁间对应的值,从而可以发现所有最小行为块和隐藏的行为关系,包括循环行为块;利用组合原理对带有重复变迁的最小行为块进行组合,得到结构行为块;组合所有行为块得到初始模型,利用已发现的隐藏行为关系修正初始模型得到更加精确的过程模型。通过实例分析和ProM的仿真实验验证了所提方法的可行性。

关键词: Petri网, 后继关系, 过程挖掘, 行为块, ProM框架

Abstract: Aiming at the limitations of the existing process mining method in discovering cyclic structures and hidden behaviors,a mining algorithm based on successor relations was proposed.A successor relations matrix was established based on the logs,the corresponding values in the matrix of the transitions were analyzed,and all the minimum behavior blocks and the hidden behavioral relationships were found,which including cyclic behavior blocks.The combination principle was used to combine the minimum behavior blocks with repeated transitions,and the behavior blocks structure was constructed.The initial process model were obtained by combining all behavior blocks,and the more precise process model through the model adjustment was obtained by discovered hidden behavior relationships.The feasibility of the proposed method was verified by case study and simulation experiment on the platform ProM.

Key words: Petri nets, successor relationship, process mining, behavior block, ProM framework

中图分类号: