Text to Process Model: Automating Process Model Creation from Text

The description and visualization of processes and underlying requirements are a core task in IT projects today. The translation of information between different forms of representation is used to include these efficiently in the different steps of a project, e.g., to describe the as-is and to-be states. However, the transformation of a process description into the visual representation of a process model is still a costly and complex task that often comes with an inevitable loss of information. One common challenge to be resolved during the process of modeling arises with the diversity of stakeholders involved. To improve the communication between stakeholders, different techniques of Natural Language Processing (NLP) are combined in this work to analyze the textual descriptions of Petri nets, in order to enable a transformation of texts into process models. The performed steps of analysis and transformation are means to ensure conformance between process models and their textual description.