【作 者】：
翟雪松， 季 爽， 李 艳
【关 键 词 】：
泛在学习； 智能体； 元宇宙； 高阶思维能力； 人机协同学习
【栏 目】：
其他
【中文摘要】：
全面构建泛在可及的终身教育体系是教育强国的重要目标，数字化赋能泛在学习模式的方式方法是教育改革与创新的关键议题。传统泛在学习在智能化、情境感知和延展方面存在痛点，难以有效促进学习者高阶思维发展。近年来，智能体与元宇宙技术的成熟为个性化内容生成和情境渲染与感知提供了技术支持，推动大规模数字化泛在学习的实施成为可能。研究构建了基于智能体+元宇宙的数字化泛在学习平台，招募118名中小学学生参与“多民族文化”主题的数字化泛在课程，通过行为序列和认知网络分析学习者交互数据。结果表明，数字化泛在学习模式显著提升了学生的问题解决能力和批判性思维，创造性思维虽有提升，但未达到显著水平。行为序列分析显示，高阶思维能力提升组在智能求解、视频学习、对话记录查看和寻求学习导引等模块间交互密切。为进一步培养创造性思维，未来研究可引入博弈机制设计多智能体，提升其在模拟情境中的自主行动能力，促进学习者以多元视角建构多人—多智能体交互关系。
【英文摘要】：
Building a comprehensive and accessible lifelong education system is a key objective in building a leading educational nation, and digital empowerment for ubiquitous learning has become a critical focus of educational reform and innovation. Traditional ubiquitous learning faces significant challenges in intelligentization, contextual awareness and scalability, limiting its effectiveness in fostering learners' higher-order thinking skills. The maturation of intelligent agents and metaverse technologies in recent years have provided technical reserves for personalized content generation and contextual rendering and perception, making it possible to promote the implementation of large-scale digital ubiquitous learning. This study developed a digital ubiquitous learning platform integrating intelligent agents and metaverse technologies. A total of 118 primary and secondary school students were recruited to participate in a digital ubiquitous course on "multi-ethnic cultures", and learner interaction data were analyzed through behavioral sequences and cognitive networks. The results show that the digital ubiquitous learning model significantly enhances students' problem-solving and critical thinking skills, while creative thinking is improved but not to a significant level. Behavioral sequence analysis indicates that the higher-order thinking enhancement group demonstrate intensive interactions among modules such as intelligent problem-solving, video-based learning, dialogue log review, and seeking learning guidance. To further cultivate creative thinking, future studies could introduce game-theoretic mechanisms to design multi-agent systems, enhancing their autonomous capabilities in simulated scenarios and enabling learners to construct multi-perspective human–multi-agent interaction frameworks.