KT4EQG: AI Framework Personalizes Exercise Questions via Knowledge Tracing
Researchers propose KT4EQG, a framework that combines educational question generation (EQG) with knowledge tracing (KT) to create personalized exercises for students. The system models each student's knowledge state from historical performance and selects the most suitable concept to practice, aiming to maximize overall knowledge mastery. An LLM-based question generator then produces targeted questions. The approach addresses the lack of fine-grained personalization in existing EQG systems.
Key facts
- KT4EQG is a personalized EQG framework.
- It uses knowledge tracing to model student knowledge states.
- The framework selects the most suitable knowledge concept for practice.
- It aims to maximize a student's potential improvement in overall knowledge mastery.
- An LLM-based question generator produces the questions.
- The approach addresses limitations of existing EQG systems.
- The paper is on arXiv with ID 2605.23933.
- The announcement type is cross.
Entities
Institutions
- arXiv