10-11, 13:30–14:00 (Europe/Berlin), Nürnberg AEG 1.01 Großer Saal
We have students with vastly different educational backgrounds, but in university settings it is often impossible to take care of all their individual needs. To mitigate this problem, we develop the ALEA system, an adaptive learning assistant. At the current stage of development, ALEA offers the following main services to the students: semantically annotated lecture notes and slides, flashcards, public comments, and private notes. The annotation uses STEX (see https://github.com/slatex/sTeX/), a semantic variant of LATEX that provides a framework for representing ontological relations between concepts. ALEA is accessible in a browser via https://courses.voll-ki.fau.de. In our talk, we present the course portal as well as the evaluation of the system. We deploy the system in different courses and evaluated it in an introductory course on Artificial Intelligence. The course was attended by about 500 students and we counted on average about 60-150 unique visits per day in our system during the lecture period. At the end of the semester, we published a voluntary survey to receive feedback for the system. In this survey, only nine persons participated but it still gives us some useful results on how to further develop the system: Out of the annotated slides, the flashcards and the lecture notes as study material, the students mostly prefer the slides, followed by the flashcards and the lecture notes. The slides are directly combined with videos from the lecture itself and a vast majority of the participants use the videos alongside the slides. Further features which the participants ask for and what we are currently developing is the inclusion of old exam tasks into the learning assistant so that students can train with them. ALEA not only offers semantically annotated learning objects but also a learner model which aims at providing students with the right material at the right time in form of a guided tour. The learner model is based on Bloom’s revised taxonomy and can thus give for each combination of learning objects and one of the cognitive processes a probability value representing the student’s ability.
Hybride Teilnahme möglich:
Webinar-ID: 653 0510 8640