A Classification Approach to Recognize On-Task Student’s Behavior for Context Aware Recommendations

Lisa Roux (1) , Thierry Nodenot (1) , Patrick Etcheverry (1) , Pantxika Dagorret (1) , Christophe Marquesuzaa (1) , Philippe Lopistéguy (1)

1 LIUPPA - Laboratoire Informatique de l'Université de Pau et des Pays de l'Adour

The increasing development of e-learning systems has raised the necessity to apply recommender systems with the aim of guiding learners through the various courses, activities, etc. at their disposal. The learner-oriented approaches allow the recommendations to fit the user's needs as precisely as possible. Nevertheless, due to the multiplicity of possible educational situations and individual particularities, offering adaptive recommendations and diversity is still a major challenge. In order to improve this aspect and provide the learner with recommendations appropriate to both their current specific needs and general profile, we focus on an hybrid system whose knowledge will be augmented through the learner's activity and results. This system will base its analyses and future recommendations according to the evolving student's profile and behaviour during the task. For that purpose, a first step is to categorize the on-task student's behaviour. This paper focuses on this problem and proposes a model, provided by educational sciences, on which the recognition process could be based.

Mots clés: Learner behaviour model; E-learning; Supervised Classification; Recommender Systems

Domaines: Sciences de l'information et de la communication; Histoire, Philosophie et Sociologie des sciences

Lisa Roux, Thierry Nodenot, Patrick Etcheverry, Pantxika Dagorret, Christophe Marquesuzaa, et al.. A Classification Approach to Recognize On-Task Student’s Behavior for Context Aware Recommendations. Intelligent Tutoring Systems, 13284, Springer International Publishing, pp.161-170, 2022, Lecture Notes in Computer Science, ⟨10.1007/978-3-031-09680-8_15⟩. ⟨hal-03705811⟩