Data mining model for better admissions in higher educational institutions (heis)—a case study of bahrain

Abstract

Data mining has been used for a variety of objectives for improving the quality of higher education institutions and especially for improving students’ performance and institution quality. The use of data mining for assessing prior learning and for improving the admission criteria has not been addressed extensively. Guiding applicants to select the correct and most suitable degree based on their prior learning at their previous institution is of great importance. We present in this paper our approach of using data mining for guiding applicants to decide the correct and most suitable degree based on their prior learning at their previous institution, and the results demonstrate the success of this method and confirm the expected benefits for the students and the institutions. The C4.5 decision tree algorithm is applied on successfully graduated student’s prior learning data along with the GPA and programme in HEI in order to predict the programme of new applicants/students of similar prior learning characteristics. The outcome of the decision tree predicted the list of appropriate programmes with the GPA expected if registered in that programme, for the applicants from similar prior learning attributes. The decision rules present a list of choices of programmes to which new students can enrol with a hint of the success level expected in terms of GPA, which gives a forecast/projection on the success level that can be expected at the end of the study tenure. Furthermore, this knowledge can be used by advisors in preventing a student from enrolling to an inappropriate programme which would make the student fail from graduating.