摘要
隐变迁是指一些存在于过程模型中,但没有出现在日志序列中的变迁。这样的变迁会大量存在于现实的模型中。从事件日志中寻找挖掘隐变迁的方法是过程挖掘技术的一个重要的难题。目前针对自由选择网有一些解决办法,但是对于复杂的过程模型有一定的局限性。本文提出了基于Petri网行为轮廓寻找隐变迁的方法。首先根据发生频率最高日志序列得出源模型,再根据剩余的日志序列一步步优化源模型从而找到隐变迁,最后通过评价指标来判定模型的合理性。
The hide transitions exist in the process models, but can not be found in the log sequence. Such tran- sitions exist in the realistic models. It is one of the important difficulties to find the methods about mining the hide transitions from the event logs. There are some solutions to the free choice nets, but they have some limita- tions due to the complex process models. The method to find hide transitions based on the behavioral profiles of Petri net is proposed in the paper. First of all, according to the highest frequency of occurrence log sequence the source model is obtained, and then, according to the remaining log sequence step by step optimization model, the hidden transitions are required. Finally, the fitness of the model by means of evaluation index is judged.
出处
《安徽理工大学学报（自然科学版）》
CAS
2016年第3期13-19,共7页
Journal of Anhui University of Science and Technology:Natural Science
基金
国家自然科学基金资助项目(61572035
61272153
61402011)
安徽省自然科学基金资助项目(1508085MF111)
安徽省高校自然科学基金资助项目(KJ2014A067
KJ2016A208)
关键词
事件日志
行为轮廓
隐变迁
过程模型
Event log
Behavioral profiles
Hide transition
Process model
