Use of Predictive Analytics for Providing Better Learning Goals in Higher Education

Abstract

Academic Mining is an interdisciplinary and developing research area of Data mining, statistics and Mathematics.  Academic Mining is a field of machine learning which deals with mining of educational data for finding out interesting patterns in various educational organizations which could help in gathering important insights. For students’ better performance, it is important to study and analyze the educational academic data. From the data mining community, the academic mining is ready to leverage an enormous amount of research and this research is applied to various educational problems including learning, evaluation assessment, and cognition.The data used in this work is a primary data collected from University of Kashmir for various students enrolled in a course. This data consists of 9 features namely students’ - Roll Number, Registration Number, College, Session, Name, Father’s Name, Subjects, Marks of subjects and Result. There are a total of 37898 entries in the dataset. The framework proposed here first extracts information from student data files, pre-processes it and then analyses it and then represents it using charts to see the performance of the students in various subjects.This can help in identifying students who need special attention in some subjects. The framework goes beyond and further analyses data for making predictions about the subject combinations in which the student can perform best. This can be used by the students to make a decision about the courses he/she can opt for higher studies.