Intelligent Educational Decision-Making System Driven by Multimodal Data Fusion and Knowledge Graphs

Abstract This paper presents a novel intelligent educational decision-making framework that combines multimodal data fusion with structured knowledge graph reasoning to enhance personalized learning experiences. To address the challenges of heterogeneousdata integration and interpretability, we propose the Cognizant Instructional Field Network (CIFNet), a hybrid neural-symbolic architecture. CIFNet integrates symbolic representations of learner states with deep contextual embeddings, supporting dynamic,interpretable decision-making processes in educational environments. It jointly models the learner’s epistemic progression,pedagogical intents, and instructional dependencies while accounting for uncertainty and sparse feedback. Building on CIFNet,we introduce the Pedagogical Inference Controller (PIC), a meta-cognitive strategic layer that further refines instructionalactions. PIC employs strategic utility estimation, regret-aware adaptation, uncertainty-weighted exploration, and curriculumalignment to optimize decision policies in real time. By simulating counterfactual instructional outcomes and prioritizingknowledge gap reduction, PIC ensures that interventions remain pedagogically coherent, learner-centered, and ethicallyaligned. Experimental evaluations across diverse educational datasets demonstrate that our integrated framework significantlyoutperforms traditional baselines and recent deep learning models in both predictive accuracy and educational outcomeoptimization. Our approach effectively bridges the gap between explainability and adaptability, paving the way for scalable,context-sensitive intelligent educational systems. The synergy between CIFNet and PIC highlights the critical role of combiningsymbolic reasoning with neural representation learning to realize robust, transparent, and adaptive educational decision-making.