This paper investigates different methodologies applied to reinforce learning techniques depending on each student's profile, which is inferred from databases generated through student interaction on the platform. The database shows what type of material the student is interested in, depending on how much time they have spent using the different materials on the platform (videos, readings, self-assessment exercises, presentations, etc.). Clustering methodologies are applied to student virtual behaviors. These allow for relating student behavior to different learning preferences and, through the recommendation of engaging activities tailored to the student's profile, are used to assist them in the learning process. These activities provide contexts for developing analytical thinking, providing feedback, and recommending personalized content through simulations. The results show that when this methodology is applied, the time spent on the platform increases, with a slight increase in student satisfaction with the platform among the students in the sample.

Keywords: Adaptive e-learning systems, learning styles, machine learning algorithms.