• 论文 •    

可解决多种复杂任务的过程挖掘算法

顾春琴,常会友,陶乾,衣杨   

  1. 1.中山大学 信息科学与技术学院,广东广州510275;2.仲恺农业工程学院 计算机科学与工程学院,广东广州510225;3.广州大学 松田学院,广东广州511370
  • 出版日期:2009-11-15 发布日期:2009-11-25

Process mining algorithm to solve diversified complex tasks

GU Chun-qin, CHANG Hui-you, TAO Qian,, YI Yang   

  1. 1.School of Information Science & Technology, Sun Yat-sen University, Guangzhou 510275, China;2.School of Computer Scinece & Engineering, Zhongkai University of Agriculture & Engineering, Guangzhou 510225, China;3.Sontan College, Guangzhou University, Guangzhou 511370, China
  • Online:2009-11-15 Published:2009-11-25

摘要: 为解决事件日志中含有循环任务、重复任务和同一任务的过程挖掘问题,进一步优化企业建模方法,改进了α算法,提出了τ算法。该算法重新定义了任务之间的关联关系,提出启发式判定规则,识别出事件日志中所包含的循环任务、重复任务和同一任务;然后对事件日志进行挖掘,提取出工作流网,并还原挖掘预处理阶段重命名的重复任务和删除的单循环任务,从而得到最终的工作流模型。最后,通过实例验证了τ算法的正确性。

关键词: 过程挖掘, 循环任务, 重复任务, 同一任务, 工作流网

Abstract: To mine process from event log with cycle tasks, repetitive tasks and same tasks, and to optimize enterprise modeling method, an improved mining algorithm called τ-algorithm was presented by improving the α-algorithm. Firstly, the associations among tasks were redefined; heuristic rules were put forward to identify cycle tasks, repetitive tasks and same tasks within event log. By mining event log, WorkFlow (WF) net was extracted. And the ultimate workflow model was obtained by restoring duplicate tasks and single cycle tasks which was renamed and deleted respectively during the period of mining preprocessing. Finally, the correctness of τ-algorithm was verified by example.

Key words: process mining, cycle tasks, repetitive tasks, same tasks, workflow net

中图分类号: