Enabling Probabilistic Process Monitoring in Non-automated Environments

Business processes are crucial for every organisation as they represent the core value generating processes. Managing business processes is important to be efficient and to compete with a globalized market. Business process monitoring is an essential means to understand and to improve working procedures. It helps detecting deviations from planned procedures and brings transparency into the state and progress of running process instances. However, without automated execution of business processes via a workflow engine, the absence of execution information hampers monitoring. Often, the automated execution of business processes is neither feasible nor desirable. However, a few monitoring points can be used, when process participants interact with IT-systems. In this paper, we propose a novel approach to business process monitoring using probabilistic estimations to fill information for missing monitoring points. The applicability of the approach is evaluated with a case study in a German university hospital.