Vietnamese educators get AI lesson-planning assistant ConnectED

Researchers from Vietnam have developed ConnectED, an AI system designed to streamline lesson planning for Vietnamese educators by aligning with national curriculum standards. The system integrates VietEduQwen, a Vietnamese educational large language model fine-tuned using supervised learning and direct preference optimization (DPO), ensuring both pedagogical accuracy and student safety. ConnectED operationalizes the ADDIE framework (Analysis, Design, Development, Implementation, Evaluation) through structured prompt templates tied to Vietnam’s Official Dispatch No. 5512/BGDDT-GDTrH, where each phase acts as both a content generation step and a teacher validation checkpoint.

Evaluation on 3,119 questions from the 2025 Vietnamese National High School Exam showed VietEduQwen achieving 87.02% accuracy—outperforming Qwen3-8B by 6.10 percentage points. Surveys of 18 teachers and 214 students highlighted strong satisfaction with curriculum alignment, lesson clarity, and usability. In real-world use, ConnectED reduced lesson preparation time from 3–4 hours to approximately 30–45 minutes, with teacher-in-the-loop reviews ensuring quality. Ablation studies confirmed that both DPO training and ADDIE-based orchestration independently contributed to system performance, underscoring the value of structured teacher oversight for practical deployment.