In recent years, students face difficulties in choosing the best content from the online distance learning of MOOCs (Massive Open Online Courses). The emerged recommendations systems to solve this problem do not identify the student's prior knowledge broadly. From this problem, the main contribution of this work is the identification and reduction of the students' knowledge gap in MOOCs. As such, in this Master's thesis, we model and analyze the MOOCs ecosystems and propose a solution for recommending parts of courses. Based on a set of three experiments, we verify that our recommendations are accurate, useful and reliable. We also present new content to fill the knowledge gap of users as the main contribution of this work to the state of the art.