Incorporating Cognitive Load and Knowledge Transfer for Multi-Domain Knowledge Tracing

Synthesis: Knowledge tracing aims to assess students' dynamic knowledge states from their learning histories, but most existing methods focus on single-domain learning despite real-world scenarios involving multiple domains simultaneously. Zhang and colleagues introduce two critical factors for multi-domain settings: cognitive load arising from managing learning across domains in temporal and knowledge dimensions, and knowledge transfer where knowledge states in one domain influence related states both within and across domains. Their proposed LT-MKT method integrates textual information from questions and their associations to bridge isolated domains, improving knowledge state assessment in multi-domain learning scenarios.

Key Findings

1. Real-world learning often involves multiple domains simultaneously, introducing cognitive load and knowledge transfer as critical factors for Knowledge Tracing.
2. Cognitive load arises from managing learning across domains in both temporal and knowledge dimensions.
3. Knowledge transfer captures how knowledge states in one domain influence related states both within and across domains.
4. LT-MKT integrates textual information from questions and their associations to bridge isolated domains.
5. The method improves students' dynamic knowledge state assessment in multi-domain learning scenarios.

Connected Concepts

• Knowledge Tracing
• Student Modeling
• Learning Analytics
• Adaptive Learning
• Personalized Learning

Connected Articles

• Kt4eqg Personalized Question Generation — KT4EQG: Personalized Exercise Question Generation via Knowledge Tracing
• Huang Interpretable Knowledge Tracing 2026 — Interpretable Knowledge Tracing
• LLM Student Modeling Memory — LLM Student Modeling and Long-Term Memory Architecture
• Neural Symbolic Knowledge Tracing — Neural-Symbolic Knowledge Tracing
• Stanbkt Bayesian Knowledge Tracing — StanBKT: Rethinking Parameter Estimation in Bayesian Knowledge Tracing

Citation

Zhang, Wang, Wu, Ding, Liu, Huang, Sha, Wang, & Liu (2026). Incorporating Cognitive Load and Knowledge Transfer for Multi-Domain Knowledge Tracing.