Monitoring students’ performance and well-being in educational settings requires a holistic approach that accounts for academic outcomes, behavioral patterns, and mental health indicators. This research proposes a multimodal AI framework for longitudinal data analysis to model and generate outputs such as performance metrics, identification of relational issues among students, and detection of mental health risks. The framework aims to organize and structure diverse data sources, including tasks completed in class and at home, test microdata of answers and results, behavioral records, classroom interactions, interaction and engagement with Learning Management Systems (LMS), teacher feedback, disciplinary incidents, and contextual attributes such as socio-economic factors and class dynamics.

The methodology integrates these multimodal data streams into a continuous monitoring system, employing interpretable machine learning models and temporal analysis to track changes over time. Context-specific configurations are guided by educational theories such as the Goal Orientation Theory [4,3] and Activity Theory [1], which provide a framework for organizing and interpreting academic and behavioral data. Applications include early identification of at-risk students, real-time interventions to improve performance and relationships, and personalized recommendations to enhance well-being. This study also explores challenges in ensuring generalizability across diverse student populations and educational environments, addressing misconceptions in Student-Based Learning (SBL) [2]. These challenges highlight the importance of contextualized and theory-driven approaches to education.

References:
[1] Carole Ames. Classrooms: Goals, structures, and student motivation. Journal of Educational Psychology, 84(3):261–271, 1993.
[2] Maria Carolina DaCosta, Tadeu da Ponte, and Marcia Moura. A case study on major teacher misconceptions on applying student-based learning in business education. In Annual Meeting of the American Educational Research Association (AERA), 2008.
[3] Carol S. Dweck. Motivational processes affecting learning. American Psychologist, 41(10):1040–1048, 1987.
[4] John G. Nicholls. Achievement motivation: Conceptions of ability, subjective experience, task choice, and performance. Psychological Review, 91(3):328–346, 1984.

Keywords: Multimodal data, student performance, well-being monitoring, student-based learning, artificial intelligence.