The rise of computational methods and AI as pillar of science, requires teaching computational thinking (computations, data science, programming, ...) at scale. The major challenge is to support heterogeneous cohorts. Interactive environments and documents, as provided by e.g. Jupyter, have proven invaluable, enabling students to progress through the teaching material at their own pace and with feedback. This is e.g. used extensively in the first years at the Faculté des Sciences d’Orsay.
The next step is to bring in adaptive teaching; that is guiding actively the student through the teaching material by providing feedback to questions such as: how am I doing in the course? what should I work on next? should I do this exercise again? am I ready to tackle chapter 5?
We are currently exploring this avenue, with an ongoing implementation for Jupyter, and a focus on ease of deployment, data locality/privacy first, frugality and explainability. The overall strategy is to upgrade the training material with lightweight annotations (prerequisites, learning objectives), to collect learning analytics to track the student progress, to feed these analytics to learner models to estimate the current student competencies, and to exploit these estimates to provide the desired feedback.
The aim of this master 2 internship is to evaluate the design and implementation -- in particular by exploiting learning analytics that have been collected to evaluate the learner models and their estimates -- to explore the design and inform future developments. It’s is hosted at the LISN lab, with a joint supervision by Nicolas Thiéry and Jill-Jênn Vie, and collaborations with research software engineer(s) and other interns.
This internship is funded by the Atlas Research Chair -- AI for Teaching and Learning (AI) at Scale -- of DataIA’s Cluster. It opens opportunities for a follow-up hire for a PhD (or Research Software Engineer) position on related topics under this chair.
Prerequisites: strong training and experience in AI (neural and/or symbolic) or data science. Strong experience in programming (Python / javascript appreciated) and software engineering (open source software development appreciated). Training or experience in educational sciences appreciated.