<![CDATA[The quick development of digital learning ecosystems after educational reform in the post-pandemic era requires an increase in intelligent monitoring systems that assess student engagement and predict academic performance. Traditional learning assessment techniques frequently have flaws when detecting early disengagement signals and initiating corrective actions for at-risk students. This research proposes an Artificial Intelligence (AI)-Driven Learning Analytics method that aims to improve student engagement monitoring and academic performance prediction in digital learning environments. A fabricated LMS-based educational dataset was used, which includes behavior analysis, engagement factors, academic factors, interaction factors, and temporal learning behavior obtained from LMSs like Moodle, Google Classroom, and Canvas. Several machine learning models, including Random Forest, XGBoost, Support Vector Machine, Artificial Neural Network, and Long Short-Term Memory (LSTM), were tested. The results revealed that the LSTM model had the best performance with an accuracy rate of 95% and a ROC-AUC value of 0.98, highlighting the importance of temporal learning behavior in educational prediction systems. Some of the essential engagement factors found to be most effective were assignment submission, quiz score, inactivity period, session length, and login number. The findings make a theoretical contribution to Artificial Intelligence in Education and Learning Analytics by combining multidimensional engagement analysis, temporal behavior modeling, and explainable AI into a unified framework. In practice, the suggested framework can aid adaptive learning, early warning, individualized intervention, and evidence-based education decisions in intelligent digital learning ecosystems.]]>
Copyrights © 2026
