DAS3H: Modeling Student Learning and Forgetting for Optimally Scheduling Distributed Practice of Skills

Benoît Choffin (1) , Fabrice Popineau (1) , Yolaine Bourda (1) , Jill-Jênn Vie (2)

1 LISN - Laboratoire Interdisciplinaire des Sciences du Numérique
2 PARIETAL - Modelling brain structure, function and variability based on high-field MRI data

Spaced repetition consists in temporally distributing exposure to an information so as to improve long-term memorization for a human learner. However, most adaptive spacing algorithms rely on expert-defined heuristics and are limited to pure memorization, e.g. foreign language learning. In this article, we propose a new student model for human skill learning and forgetting, called DAS3H. We provide empirical evidence on three real-world educational datasets that DAS3H, which incorporates both exercise-skill relationships and forgetting effect, outperforms other state-of-the-art student models that consider one or the other.

Mots clés: student modeling; adaptive spacing; memory; optimal scheduling

Domaines: Autres [stat.ML]; Machine Learning [stat.ML]; Applications [stat.AP]; Intelligence artificielle [cs.AI]; Environnements Informatiques pour l'Apprentissage Humain

Benoît Choffin, Fabrice Popineau, Yolaine Bourda, Jill-Jênn Vie. DAS3H: Modeling Student Learning and Forgetting for Optimally Scheduling Distributed Practice of Skills. JDSE 2019 - Paris-Saclay Junior Conference on Data Science and Engineering, Sep 2019, Gif-sur-Yvette, France. ⟨hal-03427048⟩