An Educational Data Mining System For Predicting And Enhancing Tertiary Students' Programming Skill.

The article focuses on developing an Educational Data Mining (EDM) system to evaluate and improve tertiary students' computer programming skills using machine learning (ML) techniques. The proposed system includes a classification module that predicts students' programming skill levels (weak, average, good, excellent) based on a real-world dataset of 1,720 samples with 36 extracted features, and a learning process module that provides tailored suggestions for skill enhancement via a knowledge-based agent. Six ML algorithms—Support Vector Machine (SVM), Naive Bayes Classifier (NBC), Decision Tree (DT), Artificial Neural Network (ANN), Random Forest (RF), and k-Nearest Neighbor (k-NN)—were trained and tested, with RF and SVM (using RBF kernel) achieving the highest classification accuracies of up to 91% and 93% (with data augmentation). The system architecture supports real-time use, enabling instructors to deliver personalized feedback and improvement plans, although the performance of the knowledge-based agent remains to be empirically validated in future work.