摘要
云端课堂作为一种优质师资常态化教学帮扶的模式,其教学过程中存储的多模态数据为全面掌握线上教师、现场教师和学生的教与学行为提供了支撑,并使云端课堂教学行为的智能识别与关联分析成为可能。在云端课堂教学过程中,系统中保存的图像、视频、音频和文本四类模态数据可作为各教学主体外在行为表征分析的来源,且其相互之间具有关联性、互补性并能彼此印证。基于这一多模态数据特征融合优势所建构的云端课堂教学行为实时检测、智能识别、关联分析与问题诊断技术方案,可实现课中实时巡课、三元教学主体行为类型预测、师生同步并发行为分析及课堂师生行为问题挖掘。以某小学三年级英语学科云端课程“Look at Me”为例,该技术方案在“教学—学习—评管—研训”四维场景中能够通过坚持以学生为中心实现双师有效协同教学,增强多向交互深度助力学生个性化深度学习,全过程实时性监督保障课堂质效精细评管,双师教学行为反思助力智能精准研训。
As a model of normalized teaching assistance provided by high-quality teachers,Online Merging Onsite(OMO)synchronous classrooms store multimodal data during the teaching process,which supports a comprehensive understanding of teaching and learning behaviors of online teachers,on-site teachers and students.This makes intelligent recognition and correlation analysis of teaching behaviors in OMO synchronous classrooms possible.In the process of OMO synchronous classrooms,image,videos,audio and text data,as four types of data saved in the system,can serve as sources for analyzing the external behaviors of various teaching entities.These data are interrelated,complementary,and mutually corroborative.Based on the advantage of integrating these multimodal data features,the technical solution for real-time detection,intelligent recognition,correlation analysis and problem diagnosis of OMO synchronous classroom teaching behaviors can achieve real-time classroom observation,prediction of behavior types of the three teaching entities,synchronous and concurrent behavior analysis between teachers and students,and mining of behavioral issues in the classroom.Taking the OMO synchronous lesson“Look at Me”for the third-grade English subject in a primary school as an example,this technical solution can achieve the following results in the four-dimensional scenario of“teaching—learning—assessing and supervising—researching and training”:adhering to a student-centered approach to achieve dual-teacher effective collaborative teaching,enhancing multi-directional interaction to deeply assist students in personalized deep learning,ensuring fine-grained assessment and management of classroom quality and effectiveness through real-time supervision throughout the entire process,and facilitating intelligent and precise researching and training through reflection on dual-teacher teaching behaviors.
作者
周德青
李惠聪
穆肃
ZHOU Deqing;LI Huicong;MU Su
出处
《现代远程教育研究》
北大核心
2025年第5期105-112,共8页
Modern Distance Education Research
基金
2024年度国家社会科学基金教育学一般项目“教师数字胜任力伴随式智能测评研究”(BCA2400502)。
关键词
云端课堂
教学行为
多模态数据
行为识别
关联分析
Project-Based Learning
Generative AI
Deepening Strategy
Instructional Design
Primary and Secondary Schools
作者简介
周德青,博士研究生,华南师范大学教育信息技术学院(广东广州510631);李惠聪,硕士,华南师范大学教育人工智能研究院(广东广州510631);通信作者:穆肃,博士,教授,博士生导师,华南师范大学教育信息技术学院,华南师范大学教育人工智能研究院(广东广州510631)。
