An Approach to Support Process Model Validation based on Text Generation

In many organizations business process modeling plays an important role for understanding, documenting and also redesigning operations. However, often the validation and the usage of process models is hampered by the fact that many domain experts or not able to understand the models in detail. In order to overcome this problem, we propose an approach for automatically generating description texts from Petri Nets. Our technique combines different techniques from linguistics and graph decomposition in order to generate accurate natural language texts. Further, we support our technique with insights from the multimedia learning theory which explicitly recommends the addressing of different channels such as the auditory and the visual channel. As a proof of concept we implemented our technique in the open source modeling tool WoPeD.