Towards an Interoperable Model-driven Automated Assessment System for Computer Science Education

In recent years, the number of computer science students has steadily risen, but the time for educators to give feedback to the students remains the same. Because of this predicament, many systems for the automatic or semi-automatic assessment of student tasks were developed. A fundamental problem of this development is that most of these assessment systems deploy a different type of data representation, which leads to a lack of interoperability between the approaches. This hinders the reuse of teaching materials that need to match the targeted system, leading to situations in which instructors need to recreate their materials. In this work, we aim to close this gap by introducing a model-driven approach called Assisted Assessment. The approach uses a technology-independent assessment model to bridge instructors' and assessment systems' technical spaces, helping instructors to transform their material for various systems. We introduce Assisted Assessment by describing the scenario of an undergraduate course for software technology and how the approach can help to manage the different available assessment systems.