Further information

In many digital modeling environments, errors in models are detected automatically and then displayed to the user in the form of descriptions or error messages. However, the learning effect for students often remains low, as they cannot fully understand the underlying problems and their causes.

The aim of this work is to develop an AI-supported concept that not only informs students about errors, but also guides them step by step in an interactive dialog to identify and correct these errors themselves. Suitable methods of AI and dialog design are to be examined and evaluated.

As part of the work, existing approaches from the fields of intelligent tutors, context-sensitive dialog systems and explainable AI are to be researched, a suitable interaction concept developed and prototypically implemented.