基于计算教育学的人工智能课堂分析框架和技术实现  

The Framework and Technical Implementation of Artificial Intelligence Classroom Analysis Based on Computational Education

作　　者：张治 程抒一 王天蓉 ZHANG Zhi;CHENG Shuyi;WANG Tianrong(Baoshan District Education Bureau,Shanghai Future Learning Research and Development Center,Shanghai 201999,China;University of Shanghai for Science and Technology,Shanghai 200093,China;Baoshan District Education Service Center,Shanghai Future Learning Research and Development Center,Shanghai 201999,China)

机构地区：[1]上海市宝山区教育局上海市未来学习研究与发展中心,上海201999 [2]上海理工大学光电信息与计算机工程学院,上海200093 [3]上海市宝山区教育事务服务中心上海市未来学习研究与发展中心,上海201999

出　　处：《开放教育研究》2024年第5期87-100,共14页Open Education Research

摘　　要：人工智能技术为基于计算的教育教学研究提供了全新场景,驱动课堂教学分析进入计算时代。本研究基于“素养导向,学为中心”的课堂价值观,形成了“主动学、深度学、互动学、高效学、多样学”五维观测视角和高质量课堂分析框架模型,建立了计算教育学视域下的人工智能课堂分析技术框架,开发了包含采集层、数据层、计算层、分析层、支架层、资源层和循证层的七大功能模块,共154项学习分析指标的人工智能课堂分析系统。在此基础上,本研究以上海市宝山区161所学校10109节课为样本开展课堂分析,构建了“微观—中观—宏观”三层次课堂改进和“诊断—反思—研究”三梯度赋能教师发展的实践路径,并形成了“理论驱动”“数据驱动”“问题驱动”三类应用模式。Artificial intelligence-enabled technologies have provided new opportunities for computation-based educational research,propelling classroom teaching analysis into the computational era.Grounded in competency-oriented and student-centered classroom values,this study developed a five-dimensional observational perspective and a high-quality classroom analysis framework,which includes"active learning,deep learning,interactive learning,efficient learning,and diverse learning."The study also established an artificial intelligence classroom analysis technical framework within computational education.It developed an artificial intelligence classroom analysis system with seven major functional modules that included data collection,data processing,computation,analysis,scaffolding,resources,and evidence inquiry,in addition to 154 learning analytics indicators.Based on these,the study conducted classroom analysis with a sample of 10,109 classes from 161 schools in Baoshan District,Shanghai,creating a three-tier"micro-meso-macro"approach to classroom improvement and a three-step"diagnosis-reflection-research"trajectory for empowering teacher development.Additionally,it has formed three types of artificial intelligence classroom analysis application models:"theory-driven,""data-driven,"and"problem-driven."

关 键 词：计算教育学 课堂分析 人工智能循证实践 教学评价 

分 类 号：G434[文化科学—教育学]