【作 者】：
祁彬斌，包昊罡，郑娅峰，李艳燕
【关 键 词 】：
智联学习环境； 学习者为中心； 云边端协同； 智能教育； 教育数字化转型
【栏 目】：
学习环境与资源
【中文摘要】：
学习环境的构建是教育数字化转型、学与教方式变革的基础。推动学习环境的改造与智能升级，构建以学习者为中心的智联学习环境，实现精准推送学习服务，是发展数字教育、建设高质量教育体系的现实需求。围绕学习环境智联计算关键问题，从学习环境多模态感知与监测、多场景学习过程记录与分析、跨场域学习场景建模与推荐、人机协同学习社群建构与支持四方面界定智联学习环境的内涵。遵循“数据处理、模型训练、智能服务”的逻辑，设计出涵盖物理环境改造与数据汇聚、数据指标与算法模型构建、精准自适应支持与智能服务的整体框架。以场景化人工智能教育应用为抓手，提出跨场域学习环境设计与评测标准、智慧学习环境计算引擎及大规模智慧教室监测平台研发、循证导向的规模化示范应用的实施路径。最后，探讨了实现智联学习环境的关键挑战，包括云边端算力基础设施、智能模型的教育可解释性、人机协同与交互体验设计、数据安全与隐私保护。
【英文摘要】：
The construction of learning environments is the foundation of the digital transformation in education and the transformation of learning and teaching methods. Promoting the renovation and intelligent upgrading of learning environments, constructing learner-centered intelligent learning environments, and realizing precise delivery of learning services are the realistic demands for the development of digital education and the construction of a high-quality education system. Focusing on the key issues of intelligent computing in learning environments, the connotation of intelligent learning environments is defined from four aspects: the multimodal perception and monitoring of learning environment, the recording and analysis of learning process in multiple scenarios, the modeling and recommendation of cross-disciplinary learning scenarios, and the construction and support of human-computer collaborative learning communities. Following the logic of "data processing, model training, and intelligent services", a framework was designed covering physical environment transformation and data aggregation, data index and algorithmic model construction, precise adaptive support and intelligent service. Taking the application of scenario-based artificial intelligence in education as a starting point, the implementation paths were proposed, which included the design and evaluation criteria for cross-disciplinary learning environments, the development of intelligent learning environment computing engines and large-scale smart classroom monitoring platforms, and the evidence-based large-scale demonstration applications. Finally, the key challenges for realizing intelligent learning environments were discussed, including cloud-edge-device computing infrastructure, the educational interpretability of intelligent models, human-computer collaboration and interaction experience design, and data security and privacy protection.