Learner Behavior Modeling Framework through Socio-Cognitive Features in eLearning Environment

Understanding learner behavior is essential for enhancing adaptive e-learning systems. This paper proposes a framework for modeling socio-cognitive learner behavior that integrates both cognitive (learning styles) and social (engagement levels) dimensions. Unlike prior approaches that focus on a single behavioral domain or rely heavily on traditional data collection methods, the proposed framework leverages interaction-based features and applies deep learning techniques for automatic learner profiling. A multilayered architecture is developed to predict learning styles and engagement levels using effective activity-based indicators within the e-learning environment. The integration of these behavioral dimensions supports the construction of dynamic learner models, enabling systems to personalize content delivery and intervention strategies. This approach provides a scalable and data-driven alternative to conventional survey-based personalization and offers educators actionable insights into learner trajectories and potential disengagement risks.