5th International Instructional Technologies & Teacher Education Symposium Logo

Investigation Of University Students' Selective Course Preferences With Association Rule Mining

Alper Bayazıt

[email protected], Yeditepe University

Servet Bayram

[email protected], Yeditepe University

Turkish Council of Higher Education declared that the instructors and supervisors have some responsibilities towards students. These responsibilities are solving problems related to university life, supervising the courses to be enrolled and making suggestions according to students’ characteristics in the selection of courses. However, in this process; some problems can arise due to reasons such as lack of communication, selection of wrong courses, lack of direction and information. The purpose of the research is to determine the tendency of the students to choose courses in order to minimize the problems and make recommendations the most relevant courses. For this purpose, data consisting of 63 rows and 34 columns and including students’ information such as course enrollment, grades of the courses taken, grade point average, university placement scores and gender were prepared. Microsoft Azure Machine Learning Studio and R statistical computing package “arules” were used. The best 250 associations meeting the support value of 0.85 were listed. The results revealed that the students who choose computer science courses are often preferred to select the courses related to computer science as elective courses. Those who choose English language tend to a second foreign language, mainly Spanish. It has been observed that these students and the students chose courses related to social sciences do not prefer technical courses. In future studies, real-time course suggestions will be integrated into student information systems by determining the course selection patterns via association rule mining.

Keywords

data mining, association rules, recommendation systems, course selection