Using educational data mining techniques to increase the prediction accuracy of student academic performance

Abstract

Purpose This paper aims to evaluate educational data mining methods to increase the predictive accuracy of student academic performance for a university course setting. Student engagement data collected in real time and over self-paced activities assisted this investigation. Design/methodology/approach Classification data mining techniques have been adapted to predict students’ academic performance. Four algorithms, Naïve Bayes, Logistic Regression, k-Nearest Neighbour and Random Forest, were used to generate predictive models. Process mining features have also been integrated to determine their effectiveness in improving the accuracy of predictions. Findings The results show that when general features derived from student activities are combined with process mining features, there is some improvement in the accuracy of the predictions. Of the four algorithms, the study finds Random Forest to be more accurate than the other three algorithms in a statistically significant way. The validation of the best-known classifier model is then tested by predicting students’ final-year academic performance for the subsequent year. Research limitations/implications The present study was limited to datasets gathered over one semester and for one course. The outcomes would be more promising if the dataset comprised more courses. Moreover, the addition of demographic information could have provided further representations of students’ performance. Future work will address some of these limitations. Originality/value The model developed from this research can provide value to institutions in making process- and data-driven predictions on students’ academic performances.

A Comparison of EDM Tools and Techniques

“…Techniques Highest accuracy appeared [43] J48 100% [46] NavieBayes(NB) 98.86% [40] X-Means 86.17% [35] Support Vector Machine (SVM) 97.98% [14] Ctree 90.37% [49] Decision Tree(DT) 67% [38] Random Forest(RF) 96.4% [34] Logistic Regression(LA) 96.98% [45] Neural Network(NN) 96% [33] K-means 98.9% [36] Rule-Based 71.3% [6] CART 98.3% [4] RepTree 61.4% [6] Iterative Dichotomiser 3(ID3) 95.9% [39] IBK 82.1% [39] Simple Logistic 93.27% [44] JRip 83.46% [35] K-Medoids 84.04%…”

“…In 2019, Ramaswami et al [38] used four data mining techniques that included Logistic Regression, Random Forest, k-Nearest Neighbour and Naïve Bayes. The authors used different techniques in order to improve the prediction accuracy of students' performance by using Python.…”

An Improved Early Student’s Academic Performance Prediction Using Deep Learning

“…Additionally, a study used two data sets i.e., the academic year (2016-2017) over a 12-week semester for 240 students, and the other one, consisting of students' assessment scores for predicting the performance [20]. Several algorithms were used such as NB, RF, K-nearest neighbor (KNN), and LR, respectively.…”

A Proposed Framework for Big Data Analytics in Higher Education

“…Discovery with Models, 687 | P a g e www.ijacsa.thesai.org according to [35], finding with frameworks usually implies the substantiated adaptation of a forecasting model throughout several environments. The principal application of this EDM class is the finding of relationships among student conduct and subjective variables in the teaching environment [32].…”

“…BD adoption and widespread use in other regions, particularly in poor countries, has been bolstered as a result of this development. In accordance with [32], the diffusion of BD gives an account of how BD goes from discovery to widespread use, and how this is aided by steps taken by service providers of important technologies necessary to enhance the resources and capacities of academic institutions. [34] Stated that countries may take advantage of the numerous BD opportunities that are accessible to them in order to gain value from the huge volumes of data that are generated and, in the long term, aid in their development.…”