TO DESIGN A DATA MINING MODEL FOR IMPROVING THE QUALITY OF HIGHER EDUCATION IN INDIA

Abstract

In the 21st century, quality has taken a back seat in the field of higher education. Deteriorating standards are mainly responsible for lack of skilled professionals. The primary reason is the gap between the academia and industry which needs an in depth study of the data available to incorporate the changes as per industry requirements and standards. newlineIn spite of the best efforts put in by the teachers, the result of students is far from satisfactory. Not all the students perform well in the examinations. The lack of quality in higher education is responsible for unemployment, increase in crime etc. newlineData Mining techniques can help solve the problem by identifying the parameters which can help in boosting the CGPA. The model has taken into consideration both the academic and non academic factors, which have impacted the performance of students. This proposed Data Mining model serves as a tool to analyze various factors which affect the quality of higher education in our country. newlineThe data was collected for the identified parameters of the model. The key aspect was that the focus was on the non academic parameters as early research focused on academic parameters. The aim was to find out as whether the overall performance of the student has anything to do with the non academic parameter. The Principal Component Analysis and Two Step Clustering methods were used for reduction of the dimensions which grouped all the parameters into some key non academic parameters. newlineLater the identified parameters were further mined to find the parameters which are of utmost significance or having large effect on the CGPA of the student. The four techniques namely Multiple linear regression, Logistic Regression, C5.0 and Bayesian Network method were applied to uncover this vital information. Finally, a comparison of the four applied techniques is done to understand the result variation. The work clearly establishes the fact that not only academic parameters but also non academic parameters affect the performance of students. newline newline