【作 者】：
付道明， 仇星月， 张 梅， 刘亚纯
【关 键 词 】：
大语言模型； 泛在学习； 应用场景； 策略； 师范生
【栏 目】：
学习环境与资源
【中文摘要】：
新一代人工智能技术正在成长为新型技术基座，为泛在学习的实现提供了坚实的技术基础。研究提出了大语言模型应用于泛在学习的三个设计原则：系统性、循环演进和开放性。基于这些原则，构建了大语言模型技术支持下的泛在教育应用模式。该模式以泛在学习的设计、实施和活动评价需求为驱动，通过与大语言模型的持续互动，在多个维度形成闭环，为教师提供全流程支持。研究通过对师范生教学技能训练的泛在学习场景进行案例分析，验证了应用模式的有效性。结果表明，该模式显著提升了学生的数字化意识和教学设计实施能力。基于研究结果，提出了三个泛在学习应用策略：构建服务导向的助学机制，推动学习空间融合化；多通道感知学习环境数据，助力学习智能泛在化；建立“人—机—物”社会性交互，实现人机深度协同。
【英文摘要】：
The emerging generation of artificial intelligence is growing into a new type of technology base, providing a solid technological foundation for the realization of ubiquitous learning. This study introduced three design principles for the application of large language models (LLMs) to ubiquitous learning: systematicity, iterative evolution, and openness. Drawing on these principles, a ubiquitous education application model supported by LLM technology was constructed. This model was driven by the needs of design, implementation and activity evaluation of ubiquitous learning, and formed various closed loops through continuous interaction with LLMs to provide teachers with the whole-process support. Through the case analysis of the ubiquitous learning scenario of teacher training in teaching skills, the validity of the application model was verified. Results indicate that this model significantly enhances students' digital awareness and their instructional design and implementation skills. In light of the research results, three application strategies for ubiquitous learning are proposed: establishing a service-oriented tutoring mechanism to promote the integration of learning spaces; leveraging multi-channel learning environment data perception to facilitate the pervasive learning intelligence; and establishing "human-computer-object" social interactions to achieve deep human-machine collaboration.