Process Mining文献阅读笔记

第一篇： 
"Prescriptive Business Process Monitoring for Recommending Next Best Actions" -Sven Weinzierl, Sebastian Dunzer, Sandra Zilker, and Martin Matzner

1. Predictive business process monitoring(PBPM) predict future process behaviour. -> To improve operational business processes.
2. PBPM constructs predictive models from historical event log data. -> Tackle different prediction tasks like: predicting next activities, process outcomes or remaining time.
3. use deep neural networks(DNNs) to learn predictive models for producing more accurate predictions in running process instances. DNN属于深度学习算法，深度学习又是machine learning的subarea. After learning, models can predict the next most likely activity of running process instances. 
4. 然而，提供下一个最可能的活动并不一定能在流程执行中支持流程的利益相关者. Organisations measure the performance of the process through key performance indicators(KPI) in regard the three dimensions: time, cost and quality. 即使event log有可能包含kpi的信息，但并不会直接影响算法的学习过程，除非a koi is the (single) learning target itself. 因此，习得的模型可以输出下一个activity的预测，这对过程的利益相关者来说是不太有利的（ 意思是算法的实际学习过程和organisation的kpi并不一致？因为利益相关者通过kpi衡量绩效，所以想要能够满足kpi绩效的next activity预测）
5. ->因为PBPM对process stakeholders less beneficial, 所以这篇论文提出了PrBPM方法。
   PrBPM评估kpi对process performance的影响，进而防止出现非organisation想要的活动。现有方法都是生成警报或recommend actions, 但是没有一个approach以process activities 的形式推荐下一个在运行流程的特定kpi方面最优的最佳action。最佳在我们的语境中指的是关于流程实例未来进程的kpi的最佳值。另外，基于kpi预测的next best actions可能会掩盖actual business process. 因此&