Predictive data driven dashboard as an academic guidance support platform for mentors

Abstract

[en] The early prediction of a student’s performance through their academic career is very relevant for mentor guidance and personalized advice. We present a combined data science and web application dashboard for analyzing real academic data from 4 degrees in the Universitat de Barcelona. Very satisfactory results are achieved on the prediction of course grades. After obtaining the predictions, We display different visualizations of the predictions in a dashboard platform to help mentors understand their set of students and identify ones in need of special attention. Big opportunities open for the future.