【作 者】：
叶芮杏， 孙福海， 黄甫全
【关 键 词 】：
AI全科教师； 机器学习； 对话代理； 视觉代理（情感计算）； 双师课堂
【栏 目】：
学习环境与资源
【中文摘要】：
随着技术的飞速发展，人工智能所驱动的技术和设备开始进入课堂。与传统教学相比，人工智能参与的这种新的协作教学形式，重新定义了学生和教师等行为者的相互角色，其所面临的主要挑战在于，需要借助技术为AI全科教师构建有意义的教学经验，从而解决教学专业技能不足、感知力欠缺和反馈性弱等问题。该研究首先采用教育学经典的知识论与认识论作为理论基础，设计了对教学极为重要的AI全科教师知识体系以及双师课堂混合智能模式。其次，基于任务导向型技术构建了问答与对话系统，基于视觉代理技术感知学习者表情特征，精确计算有关教学互动和情绪状态的多模态数据。在此基础上，拥有在复杂、非结构性以及动态环境中作出判断能力的AI全科教师便可胜任重复性的知识传授任务，与人类教师一起形成新的“双师课堂”教学模式。
【英文摘要】：
With the rapid development of technology, technologies and devices driven by artificial intelligence are beginning to enter the classroom. Compared to traditional teaching, this new form of collaborative teaching involving artificial intelligence redefines the mutual roles of actors such as students and teachers. The main challenge it faces is the need to build meaningful teaching experience for AI general teachers with the help of technology to address the problems of insufficient teaching professional skills, lack of perception and weak feedback. This study first uses the classical theory of knowledge and epistemology in education as the theoretical foundation to design a knowledge system for AI general teachers that is essential for teaching and learning, as well as a hybrid intelligence model for dual-teacher classroom. Secondly, a question-and-answer system is constructed based on task-oriented technologies, sensing learner expression features based on visual agent technology, and accurately computing multimodal data related to teaching interactions and emotional states. On this basis, AI general teachers with the ability to make judgments in complex, unstructured and dynamic environments are able to teach repetitive knowledge, forming a new "dual-teacher classroom" teaching model together with human teachers.