On the Design of a System to Predict Student's Success

Predictive models to evaluate student's performance have been widely used in the past. These models have been basically used as a statistical tool to predict whether students will pass a course based on previous background variables such as prior-learning or academic records. These models have a large potential to give support to teachers and learners during the learning process in real time. This paper focuses on the design foundations of a predictive core system. This core system is the essential component to build in the future a predictive support framework. Additionally, experimental results are shown to validate the quality of the designed system.