Hybrid Filtering Recommendation System in an Educational Context: Experiment in Higher Education in Morocco

Personalization In Oles

In a classical teaching approach, the teacher is obliged to give the same pedagogical content to a group of students, without being able to take into account the differences between them in terms of levels and preferences. This becomes possible in OLEs, by personalizing activities and content to learners, by adapting the pace of teaching, and by considering their motivations (Clément et al., 2014). Personalization also consists of proposing content and learning paths adapted to learners, by taking into account their preferences and objectives (Bejaoui et al., 2017; Garrido et al., 2016; Herath & Jayarathne, 2018).

Several research studies have examined the proposal of personalization approaches in OLEs. It is often the personalization of content and learning pathways that are discussed (Bejaoui et al., 2017; Garrido et al., 2016; Herath & Jayara-thne, 2018; Klasnja-Milicevic et al., 2011). Works in this area have tried to rely on different concepts each time to try to propose efficient systems that are well adapted to specific contexts. Some researchers have suggested hypermedia to provide learners with the opportunity to navigate in the e-learning system according to their needs (Tsortanidou et al., 2017). Other researchers have relied on intelligent tutoring systems to provide personalized learning environments (Clément et al., 2014; Segal et al., 2019). These systems use Artificial Intelligence (AI) techniques to imitate the behavior of a human tutor. Note that AI approaches are strongly used to ensure the personalization of OLEs, given their predictive capacity. Mandin et al. (2015) proposed a skills-based model of personalization of learning. Cakula & Sedleniece (2013) used the principles of Knowledge Management to present a model of personalization based on ontologies.