Designing personalized curricula based on student preferences
Periklis Georgiadis, Vassilis Christophides, Nicolas Spyratos · 2007
We address the problem of generating entire course sequences, given a set of target skills along with possibly prioritized student preferences over course descriptions. Compared to logic frameworks formulating course sequencing as a planning problem, our work relies on a set-theoretic framework for generating course sequences using preference-based queries. We introduce the concept of ordered partition for sequencing, and the ordered product of partitions, when it is necessary to combine more than one preference orderings. In our context, ordered partitions originate from preferences expressed over general relations, rather than on functional attributes of traditional database tuples (or objects) addressed by other approaches. We believe that the proposed framework is expressive enough to produce course sequences from descriptions expressed in diverse data models (e.g., XML, RDF/S) with respect to a variety of user preferences, also including priorities over the preferences.