A Simple Logic Transition Repair Method for Business Process Models via Logic Petri Nets

Abstract

Business processes are usually described and verified by process models before their realization in enterprise information systems. When event logs generated from systems cannot be replayed well in the process models, the process models need to be improved accordingly. Most model repair methods based on logic Petri nets are often designed for specific structures, which is difficult to be extended in reality. This paper proposes a simple logic transition repair method for logic Petri net-based business process models. This method can check whether models match event logs. According to different structures, we obtain the deviation positions based on the results of conformance checking. Without adding invisible transitions and repeated transitions, the model is repaired by adding arcs and logic transitions, and the original structure is maintained. Finally, the simulation experiment illustrates the correctness and effectiveness of the proposed model repair method.