In 2017, the University of Turin launched a project to help students who have fallen behind to complete their studies. The project has evolved over time and also involved CSI, leading to the application of machine learning algorithms to analyse and predict career pathways, with the aim of providing a personalised and data-driven educational model.

Help for students who are behind in their studies

The “Fuoricorso” project was set up by a group of professors from the Department of Law at the University of Turin and offers students who have fallen behind in their studies the opportunity to attend online courses on the Moodle platform, with educational material and weekly tests. The initiative involved over 1,100 students and provides a concrete example of the application of technology, which has led to a reduction in the dropout rate and an increase in the number of graduates.

AI that predicts university career pathways

In 2019, in parallel with the project to help students who had fallen behind in their studies, the Department of Law launched the “In Corso” project, involving CSI in a study (which was later developed) on the use of artificial intelligence to analyse students' academic progress. We developed a machine learning algorithm that analysed the (anonymised) academic pathways of over 15,000 students since 2005/2006, including information on the year of enrolment, obtained and missing educational credits (CFUs), exam averages, gender, type of school attended and secondary school-leaving certificate grade. Based on the information analysed, the algorithm can predict the career pathways of undergraduate students, thus providing concrete data to guide the Department's teaching policies.

Data-driven decisions

The project was so successful that it was extended to other degree courses: today the system we set up for the University of Turin analyses over 200,000 careers in 62 degree courses, across 22 departments. The university therefore has an advanced service that collects and analyses various types of information on students, in aggregate and statistical form: gender, age at enrolment, nationality, domicile, previous qualifications and corresponding grades, part-time or full-time attendance, degree programme, exam average and duration of the academic programme. Through the use of artificial intelligence, the system is able to perform a predictive analysis, classifying students based on the estimated number of years required for them to graduate. This solution enables the University to make data-driven decisions, improving the courses offered and adopting targeted support strategies.