A New Paradigm of Personalized Education Driven by Multi-Agent Collaboration
Zhonglin Zhao · Academic Journal of Sociology and Management · 2025
In order to promote the in-depth development of intelligent education, this paper proposes an intelligent collaboration framework that integrates general-purpose large language models and professional small models. By integrating the broad language comprehension ability of the general model and the fine grasp of subject knowledge of the domain-specific model, and combining the teaching mechanism in learning theory, the framework realizes a diversified knowledge construction method and supports a highly personalized and dynamically adaptive learning experience. We further explore the practical application scenarios of the framework in intelligent education, and demonstrate its potential in teaching assistance, learning guidance and knowledge transfer. With the continuous evolution of artificial intelligence technology, the collaborative system is expected to become a key engine to promote the intelligent transformation of education in the future.