Use of Data Mining For Determining Higher Education Students‘ Performance

Abstract

Current advancements in communication technologies and database technologies have made it easy for organizations to collect, store and manipulate massive amounts of data. Identifying students‘ behavior in university is a great concern to the higher education managements. An appropriate decision can be made by effectively analyzing and managing the growing volume of data. The general objective of this study is to construct a predictive model that determines the higher education students‘ performance by applying data mining techniques. The study followed the six step Hybrid methodology of Knowledge Discovery Process model such as understanding of the problem domain, understanding of the data, preparation of the data, data mining, evaluation of the discovered knowledge and use of the discovered knowledge used to achieve the goal. The study tries to understand factors affecting higher education student performance based on the data collected from St. Mary‘s University from the years 2006 up to 2009. After data preparation using data cleaning, classification algorithms such as J48 Decision tree, PART Rule induction, Naïve Bayes, Logistic regression, Support Vector Machines and Multilayer Perception Neural Network were used for all experiments due to their popularity in recent related works. The study used a dataset containing 11550 instances, 21 attributes and one outcome variable to run the experiments. WEKA 3.9.2 open source software was used as a data mining tool to implement the experiments. The study also used a 10-fold cross validation and 66% split test modes for splitting the data into training and test dataset. The result of the study showed that J48 Decision tree algorithm has registered best classification accuracy of 97.84%. The results obtained in this study are interesting and encouraging to design a model that predicts higher institution students‘ performance. Previous study field, number of common course per semester, total course per semester, year, financial source, number of supportive course per semester, were identified as the major factors affecting the student performance. In this study, an attempt was made to show the use of knowledge extracted by data mining. In the future, we recommend an automatic integration of data mining with a knowledge system so as to design an intelligent system.