Gen AI Platform for Personalized Learning Recommendations in EdTech

Current online learning platforms rely on static metrics that fail to capture real-time learner engagement and behaviour, limiting personalization effectiveness. This work proposes a Generative AI–driven personalized learning system that integrates short-term behavioural monitoring with semantic knowledge evaluation. A 35-second observation window captures interaction patterns such as response speed, activity rate, and interaction consistency. These behavioural features are combined with NLP-based semantic similarity to assess conceptual understanding. A hybrid engagement-cognition model computes a unified learner score, enabling adaptive feedback. Experimental results indicate improved engagement and performance, providing a scalable and non-intrusive alternative to traditional learning systems.