A recommendation approach based on correlation and co-occurrence within social learning network

The context of our work falls within the context of social learning networks, particularly recommendation systems. A recommendation system generally consists of proposing objects and items that meet users' needs and expectations. Within social learning, recommendation systems are of paramount importance as they guide learners in their learning path and facilitate their interactions with learning platforms. However, most recommendation systems in online learning are limited to the use of explicit feedback received from learners. In addition to explicit feedback, the new generation of recommendation systems ought to promote implicit feedbacks and actions taken by the stakeholders. In this article, we propose a hybrid recommendation system integrating all the activities carried out by the learners and combining the two notions of correlation and co-occurrence. After expounding our system, the evaluation is performed on a database outlining the interaction of employees with articles available within a Deskdrop platform. The results indicate that the performance of the hybrid approach (70%) exceeds the performance of the non-hybrid recommendation system (30%), and that the hybrid system is more consistent in terms of performance as well.