ConnectED: AI System for Vietnamese Lesson Planning and Student Learning
The recently launched AI system, ConnectED, aims to enhance the entire instructional process in Vietnamese education, covering everything from lesson design aligned with the curriculum to engaging student learning and feedback for continuous improvement. Built on the VietEduQwen model, this Vietnamese educational large language model utilizes supervised fine-tuning and direct preference optimization to ensure interactions are academically sound, pedagogically suitable, and safe for students. ConnectED applies the ADDIE framework through organized prompt templates in accordance with Official Dispatch No. 5512/BGDDT-GDTrH, with each phase acting as a generation step and a validation checkpoint for teachers. The Evaluation phase links student performance data to iterative lesson enhancements. Additionally, the system features an interactive environment for students, facilitating ongoing learning signal collection. The paper can be found on arXiv, identified as 2607.28647.
Key facts
- ConnectED is a human-centered AI system for Vietnamese education.
- It supports lesson design, interactive learning, and feedback-driven refinement.
- Built on VietEduQwen, a Vietnamese educational large language model.
- Trained via supervised fine-tuning and direct preference optimization.
- Operationalizes the ADDIE framework.
- Aligned with Official Dispatch No. 5512/BGDDT-GDTrH.
- Includes a student-facing interactive environment.
- Paper available on arXiv (2607.28647).
Entities
Institutions
- arXiv