摘要
在业务流程优化过程中,隐变迁的挖掘能够完善模型并提高流程的运行效率.目有研究主要通过事件日志中活动之间的依赖关系进行隐变迁的挖掘,但很少关注模型的结构.提出了基于块结构的隐变迁挖掘方法,利用序列编码过滤将事件日志划分为平凡序列和非平凡序列,利用α+算法得到初始模型并利用块结构层次分解,然后利用非平凡子序列匹配块结构挖掘隐变迁,融合到初始模型中并优化得到目标模型.最后通过具体实例来验证该方法的正确性和可行性.
In the business process optimization, the mining of hidden transition can improve the model and the running efficiency of the process. The hidden transition through the dependencies between activities in the event log is the main method of this study, but little attention is paid to the structure of the model. A mining method of hidden transition based on block structure of process model is proposed in this paper, the event logs using sequence coding filtering are divided into trivial and non-trivial sequences, the α+ algorithm are used to get the initial model and use the block structure hierarchical decomposition, and then use non-trivial subsequence matching block structure to mining hidden transition, fusion to the initial model and optimize the target model. Finally, the correctness and feasibility of the method are verified by a concrete example.
作者
李增
方贤文
Li Zeng;Fang Xianwen(Anhui University of Science and Technology)
出处
《哈尔滨师范大学自然科学学报》
CAS
2020年第6期1-8,共8页
Natural Science Journal of Harbin Normal University
基金
国家自然科学基金项目(61402011,61572035)
安徽省自然科学基金项目(1508085MF111,1608085QF149)
安徽理工大学研究生创新基金项目资助(2019CX2068)。
关键词
块结构
序列编码
过程模型
隐变迁
PETRI网
Block structure
Sequence coding
Process model
Hidden transition
Petri net
