Context Driven Data Mining to Classify Students of Higher Educational Institutions

Literature shows that knowledge about contextual factors associated with student time to degree and CGPA could play an important role in enabling HEIs to make more accurate and informed decisions that enhance student learning. It is also seen that such knowledge could be discovered using data mining process hidden in past data of students and used for prediction of student performance as part of the decision making process. In line with this argument in this study time to degree (total number of semesters taken to graduate) and CGPA of students were studied taking into account course difficulty and semester as contextual factors. CRISP-DM process was employed to mine student data. Results showed that classification could be used as the model for understanding about student course taking pattern, CGPA, course difficulty and semester and optimize the student time to degree in terms of the course taking pattern, course difficulty and semester to achieve best CGPA. The student data pertaining to a single programme of a single university were mined. Possible decisions in terms of student categorization based on course taking pattern, course categorization based on course difficulty, student advising and provision of learning support could be taken by using the outcomes of this research.