Educational data mining (EDM) enables personalized learning, early intervention, and curriculum adjustment. It applies machine learning and statistical techniques to analyze data from learning environments. Its key possibilities include personalized learning, early warning system, curriculum optimization, improved student models, and theory development. Despite these possibilities, EDM faces limitations in terms of Data bias and incompleteness. Algorithms may reflect systematic biases or miss key variables like motivation or home environment. Mining student data raises concern about consent, transparency and data protection under the Ethical and privacy concerns. There is also complex models which may be difficult for educators to understand or interpret. EDM often overlooks qualitative aspects like creativity, collaboration, or emotional well-being. Rodrigo (2025) explores the use of EDM in improving student models by analyzing clickstream data, quiz scores, and forum posts. He also discuss how EDM helps evaluate which teaching strategies correlate with improved outcomes. He mentioned in his research that 21st century learning requires systematic, creative, and critical thinking and educational data mining offers a suite of approaches and methods to make sense of the current and future deluge of data from all these learning systems.

Sources:

Rodrigo, M. M. T. (2025). Educational data mining: Current research and open questions. Ateneo Laboratory for the Learning Sciences. https://alls.ateneo.edu/wp-content/uploads/2025/05/final-printed.pdf