A Systematic Review and Comparative Analysis of Educational Data Mining Techniques for Student Performance Prediction

Authors: Zahira Noor Quraishi, Atul Dattarya Newase

Full-text Research PDF File:   View   |   Download

Abstract: Educational institutions increasingly rely on data-driven approaches to enhance learning outcomes and reduce academic failure. Educational Data Mining (EDM) has emerged as a prominent research field that applies data mining and machine learning techniques to analyze educational data for predicting student performance. This study presents a systematic review and comparative analysis of EDM techniques used for student performance prediction across diverse educational contexts. Following a structured review process, recent empirical studies and review papers were analyzed with respect to prediction objectives, data sources, modeling approaches, evaluation metrics, and reported outcomes. The review highlights the widespread use of traditional machine learning models such as decision trees, random forest, naïve Bayes, neural networks, and support vector machines, along with emerging methods including causal modeling, fuzzy logic, and optimization-based approaches. Comparative analysis indicates that ensemble and tree-based models often achieve high predictive accuracy, while simpler models provide greater interpretability for educational decision-making. However, the findings also reveal key limitations, including overreliance on higher education datasets, limited focus on primary education, data quality challenges, and insufficient integration of predictive models with intervention mechanisms. By synthesizing trends, strengths, and gaps, this study provides a consolidated understanding of current EDM practices and offers guidance for developing robust, explainable, and context-aware student performance prediction systems.

Keywords: Educational Data Mining, Student Performance Prediction, Machine Learning, Systematic Review, Learning Analytics

Paper Id: 7.712

Published On: 2025-01-15

Published In: Special Issue - International Conference on Engineering, Economics, Management and Applied Sciences (January 2025)

Share this