基于深度学习的智慧课堂设计框架

Design Framework of Deep-Learning-Oriented Smart Class

作为智慧教育与教学实施的主阵地,智慧课堂的建设与发展是智慧教育规模化进阶的直接驱动力。然而,智慧课堂的概念庞杂且分散,其建设陷入设备堆砌和功能叠加的单向循环,并涌现诸多功能障碍。本研究通过梳理智慧课堂的研究现状,厘清其囊括外在环境条件和内在目标模式的宏观概念、深度学习的价值导向;基于智能技术赋能的智慧课堂新形态,归纳得出面向特色空间的沉浸式课堂和面向基础空间的交互式课堂应用,明晰网络与技术、数据与模型、硬件终端设备与软件系统平台、数字教学资源、教师资源五大结构要素及其功能边界;结合DIKW模型,设计包含知识组织与整合、知识获取转换与输出、知识进化的深度学习框架。该框架可用于指导智慧课堂的有序建设,辅以人机协同底线思维推动智慧课堂的高效运转,助力深度学习的有效达成与智慧生成。

As the direct product of smart education and the practice field,smart class development is a driving force for the large-scale advancement of smart education.However,the theoretical concepts of smart class are numerous and scattered,whose construction falls into a one-way circle of stacking equipment and superimposing function,and many functional barriers have emerged.According to the current smart class research,this study clarifies its macro concept of the external environment and the value orientation of deep learning.Based on the new form of smart class,we summarize the application of immersive class oriented to characteristic space and interactive class oriented to basic space and identify the structural elements and functional boundaries of five aspects,including network and technology,data and model,hardware terminal equipment and software system platform,digital teaching resources and teacher resources.Combined with DIKW,a deep learning framework consisting of knowledge organization and integration,knowledge acquisition,transformation and output,and knowledge evolution is designed.The framework will guide smart class construction in order,promote the efficient development of smart class based on a human-machine collaboration mechanism,and help to achieve effective deep learning and wisdom generation.