Hanoi (VNA) – Artificial intelligence (AI) is emerging as a practical solution to one of the biggest challenges facing personalised learning in Vietnam – delivering tailored instruction in classrooms, experts said at a seminar in Hanoi on June 4.
The seminar, titled "Personalising Mathematics Learning with AI: A Strategic Direction for Vietnamese Education," was organised by the Education Journal under the Ministry of Education and Training.
The discussion comes as the Politburo's Resolution No. 71 on breakthrough development in education and training calls for the stronger application of digital technology and AI in teaching and learning.
Technology helps overcome bottlenecks
Associate Prof. Dr Nguyen Chi Thanh, Dean of the Faculty of Education at the University of Education under Vietnam National University, Hanoi, said differentiated instruction requires teachers to adjust learning content, teaching methods, and classroom environments to meet different needs.
He explained that personalised learning is built around an individual competency profile for each student, identifying strengths, areas for improvement, learning styles and interests. Continuous assessment enables teachers to update those profiles and adapt instruction accordingly.
Associate Prof. Dr. Chu Cam Tho, Head of the VMathAI project and former Director of the Educational Assessment Research Division under the Vietnam National Institute of Educational Sciences, said adaptive learning differs from conventional teaching because every student follows a personalised learning pathway.
However, she noted that this approach faces practical challenges. Vietnam's national curriculum requires students to achieve common learning standards. Classrooms of 40 to 50 students make it difficult for teachers to understand each learner's abilities and provide differentiated instruction.
According to Tho, AI offers an effective way to address these constraints.
Unlike traditional classrooms where all students progress at the same pace, AI-supported learning platforms can analyse learning data, practice results and students' responses to recommend the most suitable next lessons. The aim is to keep learners within their "zone of proximal development" by providing tasks that are challenging enough to stimulate thinking without becoming overwhelming.
She cited VMathAI, a mathematics learning platform developed by her research team after 24 years of studying personalised education, as a practical example.
Designed as an adaptive learning solution, VMathAI combines AI with standardised digital mathematics resources to address two key questions: how to build an individual competency development pathway for each student and how to guide each learner through a curriculum suited to their abilities.
The platform has been introduced at 21 upper secondary schools, covering more than 800 classes, over 37,000 students and more than 350 teachers and education managers. More than 80% of users have expressed satisfaction with the platform, according to Tho.
She added that the model could be expanded beyond mathematics to other school subjects.
An inevitable trend
Speaking at the seminar, Bach Dang Khoa, Deputy Director of the Department of Education and Training of Bac Ninh province, said personalised and adaptive education is becoming an inevitable trend.
He stressed that teachers need to understand how well students have mastered previous lessons before preparing subsequent classes so that classroom teaching can be more effective.
"Teachers need to identify each student's strengths and competencies, and AI can help measure those individual capabilities," he said, noting that this was one reason Bac Ninh joined the VMathAI pilot programme.
Nguyen Thi Huong, Head of the Primary Education Division under Ninh Binh province’s Department of Education and Training, agreed that applying AI to adaptive and personalised learning is both necessary and increasingly urgent.
Huong said AI could also reduce subjectivity in teachers' assessment of students' abilities. However, she cautioned that AI-supported learning must remain aligned with the national curriculum, with appropriate safeguards to ensure learning tasks do not exceed expected competency levels. Ethical issues related to AI use should also receive special attention./.
