基于自编码器的带隐变迁的迹变化挖掘

Trace Change Mining with Hidden Transitions Based on Autoencoder

针对带有隐变迁的部分迹变化无法利用活动间的行为轮廓关系检测的问题,引入了一种基于自编码器的迹变化挖掘方法。业务系统参考模型未知,对原始日志中的迹执行变化操作时,添加隐变迁等操作会导致活动间的行为轮廓关系未改变、而实际迹执行存在变化。利用自编码器学习原始日志的分布特征,挖掘实际日志中的每条迹是否发生删除、插入和移动变化操作,从迹的角度出发实现活动变化挖掘。通过Pytorch仿真实验,本所提方法可以检测出实际日志中执行变化操作的迹,实验结果表明自编码器在迹变化挖掘中有效。

In order to solve the problem that some trace change operations with hidden transitions cannot be detected by using the behavior contour relation between activities, a trace change mining method based on autoencoder is introduced.When the reference model of the business system is unknown and the trace in the original log is changed, the behavior contour relation between activities is not changed by adding hidden transitions and other operations, then the autoencoder is used to learn the distribution characteristics of the original log. Mining whether Delete, Insert and Move changes occur in each trace in the actual log to realize active change mining from the trace perspective. The Pytorch simulation results show that the proposed method can detect trace with change operation in real log, and the experimental results show that the autoencoder is effective in trace change mining.