An Early Detection System for Incomplete Application in Master’s Degree program at University level

Thumbnail Image
Journal Title
Journal ISSN
Volume Title
The British University in Dubai (BUiD)
Selecting a university to pursue higher education is a very difficult and expensive decision that a student can make. A lot of research into student decision-making in this area has been conducted, but the work has been focusing mostly on students’ dropout rate after registrations or before completion of the course using surveys, general market trend and also based on other existing research work on this area. This paper address the major issue of student incomplete application rate that is faced by a higher education institute based in Dubai. By analyzing the past year student dropout records from the university database, this research intends to build a Incomplete Application Prediction Model (IAPM) and in turn come up with a strategy that can help university to achieve application completion rate. Early identification of these incomplete applications is as important or in this case can be considered more important than student marketing from the university point of view as retaining prospective students who already applied are easier if detected early as details of the students are already available. Moreover in the university for which the study is undertaken, records show an alarming rate of around 16.25% student incomplete applications after enrolling which urges the need for this study. Various data classification techniques as well as association rules were applied on all the attributes and also on selective attributes that were obtained from the university’s original database for research purpose. An Incomplete Application Prediction Model is developed from these techniques to aid student retention for master’s degree courses in a specific university. This model can be further customized for other universities in the region if needed. The results were positive and indicated that past incomplete students’ records can be a valuable resource for mining the near accurate reason for students’ incomplete application, which in turn can give the management a clear insight for proactive solving of such issues in future intake. Data visualization in Weka offered interesting insights on these data also. By focusing on antecedents of incomplete students’ records, colleges can restructure their strategies for a better student-supportive system. Smaller sample size and the self-explicated data are some limitations of this research work.
master’s degree, incomplete applications, Incomplete Application Prediction Model, student retention strategy, data classification, association rule