Meike Ullrich

Graphical modeling plays an important role in many areas of computer science. Graphical models are used, for example, in business process modeling (Petri nets, EPK, BPMN) or in software development (ER diagrams, UML) to clearly represent complex issues. In the KEA-Mod project, an e-assessment platform was developed that allows tasks for creating such models to be configured and automatically evaluated with the help of assessment services. Possible final theses on this topic could deal with the following questions, for example:

• How can an evaluation service be designed that does not directly perform the complete correction of a task, but guides users step by step to the correct solution during processing?
• Can the use of generative AI with Large Language Models in an evaluation service support the evaluation and creation of feedback messages (LLM as a judge)?
• 
  

The (new) development of evaluation services (based on an existing framework) for certain modeling languages (e.g. CMMN, DMN, DRD, Petri nets, BPMN, EPK) is also possible within the scope of this topic.