Preliminary results from a machine learning based approach to the assessment of student learning

We describe a possible approach to the problem of extracting knowledge from the analysis of questionnaires through machine learning. The idea guiding our research was to investigate the existence of association rules among the topics covered in a course. The data used came from the questionnaires administered to the freshmen in electronic engineering attending the course of foundation of computer science at our university. Each questionnaire was coded into feature vectors that were classified with respect to the grade obtained by the student and analysed with C4.5. Some statistical results and hints for further work are discussed.