Educational Data Mining
Overview
Educational Data Mining uses data from learning management systems, intelligent tutors, assessments, classroom interactions, and learner-produced work to study how people learn and where they need support. Common approaches include predicting performance or disengagement from activity logs, knowledge tracing and cognitive diagnosis, graph-based recommendation, clustering learner behaviors, and language or multimodal analysis of dialogue and classroom activity. These methods can inform personalized practice, early warnings, tutoring, and instructional redesign, while large language models are increasingly used to analyze educational content and interactions. A central challenge is ensuring that models generalize across courses and learner groups, distinguish engagement from mastery, avoid data leakage, and support privacy and fair decisions rather than treating predictive accuracy as evidence of learning gains.
Latest papers
74 papers7 papers in the last four weeks, down 22% on the four weeks before. 0.0% of all new papers.