Multi-Objective Optimization for Personalized Educational Pathway Planning: A Temporal Student Modeling Approach

Abstract

Personalized educational pathway planning requires learning systems to continuously adapt instructional sequences to the evolving knowledge, behaviour, learning objectives, and cognitive characteristics of individual students. Conventional recommendation approaches frequently rely on static learner profiles or optimize a single criterion, limiting their ability to accommodate changes in student knowledge and competing educational requirements. This paper proposes a temporal student modeling approach for multi-objective optimization of personalized educational pathways. The proposed framework models learner knowledge as a dynamically evolving state derived from sequential assessments, learning-resource interactions, progression history, engagement patterns, and temporal behaviour. A multi-objective optimization layer subsequently generates feasible learning pathways by simultaneously considering knowledge mastery, learning efficiency, knowledge coverage, cognitive difficulty, learner engagement, and prerequisite consistency. Rather than producing an immutable sequence, the framework continuously updates and re-optimizes the remaining pathway as new evidence regarding student performance becomes available. The approach integrates temporal knowledge tracing, educational dependency modeling, and Pareto-based pathway optimization into a unified framework. The proposed formulation provides a foundation for adaptive and learner-centered educational systems capable of balancing short-term learning requirements with long-term educational goals while maintaining pedagogical coherence and pathway flexibility.