Research Article
AI-Powered Real-Time Student Monitoring and Adaptive Alert System for Virtual Learning Environments
@INPROCEEDINGS{10.4108/eai.28-4-2025.2357781, author={Godasu Deekshitha and Sabolu Sharon and Sikkonda Kuruva Mamatha and Patil Varala Jyothsna and K Lakshmi}, title={AI-Powered Real-Time Student Monitoring and Adaptive Alert System for Virtual Learning Environments}, proceedings={Proceedings of the 4th International Conference on Information Technology, Civil Innovation, Science, and Management, ICITSM 2025, 28-29 April 2025, Tiruchengode, Tamil Nadu, India, Part I}, publisher={EAI}, proceedings_a={ICITSM PART I}, year={2025}, month={10}, keywords={real-time student monitoring; virtual learning environments; adaptive alert system; multimodal fusion; personalized engagement modeling; edge ai in education}, doi={10.4108/eai.28-4-2025.2357781} }- Godasu Deekshitha
Sabolu Sharon
Sikkonda Kuruva Mamatha
Patil Varala Jyothsna
K Lakshmi
Year: 2025
AI-Powered Real-Time Student Monitoring and Adaptive Alert System for Virtual Learning Environments
ICITSM PART I
EAI
DOI: 10.4108/eai.28-4-2025.2357781
Abstract
The issue that online education presents is how to ensure student engagement and provide timely instructional support in that shift. We propose PRISM-AI (Personalized Real time Intelligent Student Monitoring with Adaptive Interventions), an innovative system, which monitors student in virtual learning environment with real time multi modal AI. The scores of the engagements are obtained by fusing the data from visual, audio, and behavior domain and are dynamically compared with the personalized baselines using the Bayesian modelling. PRISM-AI triggers context aware interventions triggered when disengagement is detected based on individual’s learning needs. Unlike traditional rule-based or single-modality systems, PRISM-AI offers superior accuracy, faster response times, and adaptive intelligence. We demonstrate the effectiveness of our approach by extensive experimental results that lead to improvement in the accuracy of engagement detection (up to 91%) and its prediction reliability, as well as efficiency in interventions. Because the system is edge deployable, privacy preserving, and highly scalable, it is a practical solution for next generation online education platforms.
