Repairing Process Models Containing Choice Structures via Logic Petri Nets

The process knowledge can be extracted based on the process mining technology from event logs, which can be generated from information systems. The event logs can be mined to construct a process model. The business process recognized by information systems can be described accurately by repairing the models. Some model-consistent metrics cannot be enhanced by the existing model repair approaches efficiently, such as generalization, precision, simplicity, and fitness. Thus, in this paper, we propose an approach for repairing the models via a logic Petri net (LPN). First, it builds process models by LPN. Next, for process models containing choice structures, the model repair approaches are proposed. Specifically, some relations between the transitions in the choice structure are studied in order to decide the positions where to repair the model. Finally, some examples of thoracic surgery processes in a hospital are given. Comparing with the state-of-the-art approaches in the literature, experimental results show that the fitness and precision of the models can be improved based on our proposed approach effectively.