Utilizing educational data mining for enhanced student performance analysis in Malaysian STEM education

Educational Data Mining (EDM) applies data mining in education, aiding schools to enhance student learning programs by analyzing data and success factors. In the era of big data, schools must adopt data-driven approaches. However, predicting success among diverse secondary students in Malaysia remains uncertain due to dataset size and heterogeneity. This study aims to identify key predictor variables for STEM student performance and present a systematic method for analysis, benefiting academics, schools, and the education ministry. The article explores data mining via knowledge discovery (KDD) and employs classifiers like Random Forest, PART, J48, and Naive Bayes on a dataset of Malaysian upper-secondary Science students. Utilizing WEKA for analysis, the research utilizes 21 features from the Education Repository and SAPS. Notably, J48 outperforms other classifiers. The study aids educational enhancement, enabling early intervention and improved academic achievement.

Text 108442.pdf - Published Version Download (633kB) Official URL or Download Paper: https://ijarped.com/index.php/journal/article/view...

Actions (login required)

View Item