Prediction of academic success through interaction with version control systems

Abstract

Version Control Systems are commonly used by Information and communication technology professionals. They allow monitoring the single activity of different people working in the same project through all the project's lifetime. Thus, Version Control Systems are also used by educational institutions. The aim of this work is to demonstrate that the academic success in Computer Sciences subjects may be predicted by monitoring students' interaction with a Version Control System. In order to do so, we have built a prediction model by using Model Evaluator, a tool which allows evaluating several Machine Learning models in order to select the most suitable for a specific problem. A set of experiments are being been carried out to evaluate the prediction model. At those experiments, students of different subjects of the degree in Computer Science of the University of Leon were involved. All the selected subjects: Operating Systems Extension, Computer Programming, and Computer Organization, require students to complete practical assignments which imply developing software solutions in different programming languages. A common requisite of those practical assignments is the use of a Git repository to store source code and documentation. Tentative results suggest that continuous interaction with the Git repositories has a big impact on academic success.