Extraction of the Characteristic Attributes of Student Athletes using a Questionnaire using the Support Vector Machine

Abstract

Providing special support for student athletes and creating future educational strategies are major issues for many universities. In this study, we created a questionnaire that comprised 77 multiple choice questions. After collecting the responses from 100 student athletes and 141 other students, we analyzed the characteristic attributes of student athletes. We considered the 278 combinations of question items and the response choices as the attributes to represent the students. A student is represented as a 278 dimensional boolean vector. Then, we applied the support vector machine (SVM) machine learning method with feature selection. The result confirmed that it is possible to distinguish between student athletes and other students with a 90% accuracy based on 16 characteristic attributes of the students, such as the following: they spend a lot of time in athletics clubs and not on their studies; they want to work for an economically rich life; they think that it is advantageous to search for jobs or complete their graduation if they have good grades; they have less interests in the international activities in campus life. Students aiming for professional and semiprofessional athletic careers are motivated for studying because of the competitiveness cultivated in their athletic careers.