Abstract:

This paper presents the first findings and products developed within the European ERASMUS+ project SPEET (Student Profile for Enhancing Engineering Tutoring).

This project emerges from the potential synergy among:
a) the huge amount of academic data actually existing at the academic departments of faculties and schools, and
b) the maturity of data science in order to provide algorithms and tools to analyse and extract information from what is more commonly referred to Big Data. A rich picture can be extracted from this data if conveniently processed.

The main purpose of this project was to apply data mining algorithms to process this data in order to extract information about and to identify student performance characteristics. The main rationale behind this proposal is the observation that students performance can be classified according to their behaviour while conducting their studies. After years of teaching and sharing thoughts among colleagues from different EU institutions it seems students could obey to some classification according to the way they face their studies. Therefore, if it would be possible to know what kind of student one student is, this may be of valuable help for tutors.

Therefore, to determine and categorize the different profiles for engineering students across Europe. In fact, at this moment this categorisation has been conducted at the countries that participate of the project, with the aim of extending the study.
Specific tools have been developed to analyse available student data and to elaborate student clusters depending on the expected student performance. What is important here is the possibility to conduct explanatory analysis in order to understand what are the main features that determines one student belongs to one cluster. In such way, for a new student, it would be possible to know in advance the possibility to belong to one cluster or another on the basis of its characteristics (sex, admission age, previous studies, first semester marks, etc). This potential is proposed here to be redirected to help tutor’s to better know their students and improve counselling actions.

This main use of this student profile analysis is that of being embedded on supporting IT tools for tutoring. Once key labels for the different profiles are determined, there will be the need to determine the profile one student complies with as it starts. The first results along with collateral data should allow the IT tool identify the student’s profile (or potential profiles if doubts) and help the tutor to know how to provide the student with the appropriate suggestions in order to increase it’s performance and satisfaction with the studies. An immediate step further is that of extending the analysis to other disciplines than engineering (social sciences, medicine, etc) and compare (if any difference) the students profile that arise. The comparison can be done country and discipline wise.

A series of engineering degrees from five different European countries have been analysed. Data from the considered institutions has been processed and student profiles characterized. Some common features arise from the different countries.

Keywords:

International projects, New Experiences for the International Cooperation, Educational Data Mining.