计算机工程与应用 ›› 2017, Vol. 53 ›› Issue (20): 237-242.DOI: 10.3778/j.issn.1002-8331.1703-0093

• 工程与应用 • 上一篇    下一篇

融合特征网与模块网的业务过程挖掘

程腾腾,方贤文,王丽丽,刘祥伟   

  1. 安徽理工大学 信息与计算科学系,安徽 淮南 232001
  • 出版日期:2017-10-15 发布日期:2017-10-31

Business process mining based on feature nets and module nets

CHENG Tengteng, FANG Xianwen, WANG Lili, LIU Xiangwei   

  1. Department of Information and Computing Science, Anhui University of Science and Technology, Huainan, Anhui 232001, China
  • Online:2017-10-15 Published:2017-10-31

摘要: 过程挖掘目的是通过分析由信息系统记录的日志得出的过程模型,从而改善和维护业务流程。目前,许多业务流程都以模块化的方式进行交互。虽然很多过程挖掘算法已经被提出来,不过对于处理多模块还有一定的局限性。提出了基于特征网与模块网的挖掘算法,根据日志将特征分为不同模块;在此基础上,分别求出模块间特征交互的特征网与模块内的特征交互模块网;将两者根据提出的融合算法进行融合,得到完整的过程模型。通过一个用户网上购物的实例说明了该算法的可行性。

关键词: Petri网, 过程挖掘, 过程发现, 业务流程

Abstract: Process mining techniques aim to improve and maintain the business process via analyzing process model, which is derived based on the event log of information system. At present, many business processes interact with each other in a modular way. A lot of algorithms have been proposed, the processing of multi module still has limitations. A business mining method is proposed based on feature nets and module nets. According to the log, features are divided into different modules. On this basis, a feature net with features interactive between module and modular nets with features interactive in the module will be gained. Then modular nets and feature nets are fused, and a complete model will be obtained. Finally, the feasibility of the algorithm is illustrated through an example of online shopping.

Key words: Petri net, process mining, process discovery, business process