Deep and Surface Representation of Competences in Academic Curricula

The curricular design of programs in Higher Education is increasingly oriented towards competence-based learning goals in order to provide more specific and explicit qualifications. At the same time the descriptions of competences are mostly based on textual descriptions often in informal style targeted towards different audiences making descriptions vague and harder to compare. Formal models for the specification of competences exist for some time and even semantic models are being developed, e.g., the European Learning Model. These different approaches lead to a gap between formal and informal competence descriptions that require manual efforts of curriculum designer to maintain. This research borrows the concept of Deep and Surface representation models from the field of linguistics to unify the different approaches. Additionally, it focuses on the functionality of Generative Artificial Intelligence in the form of Large Language Models to close the aforementioned gap. It demonstrates the potential of the technology in both directions and discusses application potential of the concept.