Comparing Machine Learning Models for Classifying Online Learning Satisfaction: A Supervised and Unsupervised Approach

The rapid expansion of online education has made learner satisfaction a central measure of platform quality, yet many institutions still rely on end-of-course surveys that capture feedback too late to support timely intervention. This study examines whether supervised and unsupervised machine learning can predict and explain student satisfaction in online learning using real learner feedback. The publicly available Online Education System Review dataset from Kaggle, comprising 1,033 learner records, was used as the basis for the analysis. Four supervised classifiers, namely Random Forest, Naive Bayes, Logistic Regression, and K-Nearest Neighbours, were trained in WEKA to classify satisfaction into three levels, while Simple K-Means clustering was applied to uncover natural learner segments. Random Forest delivered the strongest and most balanced performance, with an accuracy of approximately 63.5 percent and a weighted ROC area of 0.757, followed closely by Naive Bayes and Logistic Regression, whereas K-Nearest Neighbours performed noticeably worse at about 51.6 percent. Across the supervised models, instructor support, interaction frequency, content quality, and platform usability emerged as the most influential predictors of satisfaction. Clustering revealed distinct learner groups that differed in engagement and demographic profile while sharing comparable satisfaction tendencies. The results indicate that ensemble classification combined with clustering provides a practical, interpretable, and reproducible approach for monitoring satisfaction and informing course design. The study contributes to learning analytics by showing how accessible tools can convert routine learner feedback into evidence for improving digital education, while acknowledging the limits of single-snapshot survey data

online learning satisfaction; machine learning; learning analytics; Random Forest; predictive modelling

1. Addas, A., Naseer, F., Tahir, M., & Muhammad, N. K. (2024). Enhancing higher-education governance through telepresence robots and gamification: Strategies for sustainable practices in the AI-driven digital era. Education Sciences, 14(12), 1324. https://doi.org/10.3390/educsci14121324 [Google Scholar] [Crossref]

2. Cinar, O. E., Rafferty, K., Cutting, D., & Wang, H. (2024). AI-powered VR for enhanced learning compared to traditional methods. Electronics, 13(23), 4787. https://doi.org/10.3390/electronics13234787 [Google Scholar] [Crossref]

3. Covrig, M., Goia, S. I., Igret, R. S., Marinas, C. V., Miron, A. D., & Roman, M. (2023). Students’ engagement and motivation in gamified learning. Amfiteatru Economic, 25(Special Issue 17), 1003–1023. https://doi.org/10.24818/EA/2023/S17/1003 [Google Scholar] [Crossref]

4. Deng, P., Chen, B., & Wang, L. (2023). Predicting students’ continued intention to use e-learning platform for college English study: The mediating effect of e-satisfaction and habit. Frontiers in Psychology, 14, Article 1182980. https://doi.org/10.3389/fpsyg.2023.1182980 [Google Scholar] [Crossref]

5. Desai, S. P., Munshi, M. M., Hanji, S. V., & Pabba, C. (2024). Impact of the learning environmental factors on group engagement: A deep learning approach. South Asian Journal of Management, 31(6), 150–173. https://doi.org/10.62206/sajm.31.6.2024.150-173 [Google Scholar] [Crossref]

6. Hassan, B., Muhammad, O. R., Siddiqi, Y., Muhammad, F. W., & Rabiya, A. S. (2025). CONNECT: An AI-powered solution for student authentication and engagement in cross-cultural digital learning environments. Computers, 14(3), 77. https://doi.org/10.3390/computers14030077 [Google Scholar] [Crossref]

7. Joseph, A. R. (2024). The impact of personalized learning on intrinsic motivation, academic proficiency, and behavioral issues [Doctoral dissertation. ProQuest Dissertations and Theses (Order, (31146874)). [Google Scholar] [Crossref]

8. Lampropoulos, G., & Sidiropoulos, A. (2024). Impact of gamification on students’ learning outcomes and academic performance: A longitudinal study comparing online, traditional, and gamified learning. Education Sciences, 14(4), 367. https://doi.org/10.3390/educsci14040367 [Google Scholar] [Crossref]

9. Naseer, F., Muhammad, N. K., Addas, A., Awais, Q., & Ayub, N. (2025). Game mechanics and artificial intelligence personalization: A framework for adaptive learning systems. Education Sciences, 15(3), 301. https://doi.org/10.3390/educsci15030301 [Google Scholar] [Crossref]

10. Sujaradha, G. (2021). Online education system review [Data set. Kaggle. https://www.kaggle.com/datasets/sujaradha/online-education-system-review [Google Scholar] [Crossref]

11. Sulisworo, D., Wulandari, Y., Effendi, M. S., & Alias, M. (2021). Exploring the online learning response to predict students’ satisfaction. Journal of Physics: Conference Series, 012117. https://doi.org/10.1088/1742-6596/1783/1/012117 [Google Scholar] [Crossref]

12. Tu, Y., Chen, W., & Brinton, C. G. (2020). A deep learning approach to behavior-based learner modeling. https://doi.org/10.48550/arXiv.2001.08328 [Google Scholar] [Crossref]

13. Yeganeh, L. N., Nicole, S. F., Chen, Y., Simpson, A., & Hatami, M. (2025). The future of education: A multi-layered metaverse classroom model for immersive and inclusive learning. Future Internet, 17(2), 63. https://doi.org/10.3390/fi17020063 [Google Scholar] [Crossref]

14. Yuan, J., Qiu, X., Wu, J., Guo, J., Li, W., & Wang, Y. (2024). Integrating behavior analysis with machine learning to predict online learning performance: A scientometric review and empirical study. https://doi.org/10.48550/arXiv.2406.11847 [Google Scholar] [Crossref]

• The Role of Artificial Intelligence in Revolutionizing Library Services in Nairobi: Ethical Implications and Future Trends in User Interaction
• ESPYREAL: A Mobile Based Multi-Currency Identifier for Visually Impaired Individuals Using Convolutional Neural Network
• Comparative Analysis of AI-Driven IoT-Based Smart Agriculture Platforms with Blockchain-Enabled Marketplaces
• AI-Based Dish Recommender System for Reducing Fruit Waste through Spoilage Detection and Ripeness Assessment
• SEA-TALK: An AI-Powered Voice Translator and Southeast Asian Dialects Recognition