Analyzing students' academic performance using educational data mining

• Published in: Computers and education. Artificial intelligence Vol. 7; p. 100263
• Main Authors: Sarker, Sazol, Paul, Mahit Kumar, Thasin, Sheikh Tasnimul Hasan, Hasan, Md. Al Mehedi
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.12.2024; Elsevier
• Subjects: Academic performance; Classification; Consistent performance; Educational data mining; Performance progression; Educational data mining; Performance progression; Consistent performance; Classification; Academic performance
• ISSN: 2666-920X; 2666-920X
• DOI: 10.1016/j.caeai.2024.100263

Educational Data Mining (EDM) is the process of extracting useful information and knowledge from educational data. EDM identifies patterns and trends from educational data, which can be used to improve academic curriculum, teaching and assessment methods, and students' academic performance. Thus, this study uses EDM techniques to analyze the performance of higher secondary students in Bangladesh. Three crucial categories, such as good, average, and poorly-performing students are considered for analysis. Four significant aspects of students' performance are emphasized for evaluation in this study. Firstly, predicting students' academic final examination performance in terms of internal college examination. Secondly, identifying all subjects' impact on classifier performance. Thirdly, examining students' performance progression during their studies and relating with subject-wise improvement or degradation. Fourthly, discovering consistent patterns of students' performance based on previous internal examination performance trends. The classification result reveals the correlation between internal examination and final academic performance. In addition, it resembles the predictor subjects for academic performance. The result also highlights the consistent pattern of students' consecutive internal examinations' performance. Thereafter, college administration can take necessary supportive initiatives for poorly-performing students and encourage good-performer students to continue excelling. •Proposed two reasonable techniques for calculation of HSC GPA in pandemic.•Proposed GPA techniques relate more to students previous college performance.•Efficient measurement of subjects and students performance relationship.•Analyzed subject-wise performance progression for internal examinations.•Finding performance booster, degrader, and at-risk subjects.