基于表情向量聚类的课堂交互学生推荐方法
A Recommendation Method for Interactive Students in Class Based on Expression Vector Clustering
李慧 1张运良 2陈红倩 3李英侠4
作者信息
- 1. 中国科学技术信息研究所,北京 100038;北京联合大学 管理学院,北京 100101
- 2. 中国科学技术信息研究所,北京 100038
- 3. 北京工商大学 计算机学院,北京 100048
- 4. 北京联合大学 管理学院,北京 100101
- 折叠
摘要
为了提升课堂提问环节交互学生的选择针对性,提出一种基于表情序列向量聚类的课堂交互学生个体推荐方法.首先,使用表情识别模型从学生端视频中获取表情向量,将每一名学生的所有表情向量汇总至教师端,并组合为表情向量序列;其次,基于学生表情向量序列提出基于短期表情向量聚类的学生分类方法和基于长期表情向量聚类的学生推荐方法;最后,通过可视化和表格形式为教师呈现学生的分类和推荐结果,为教师快速选择不同类型中最突出的学生个体提供支持.通过教师评价发现,所提方法能有效表现学生的状态差异,可显著提升课堂交互中学生的判别效率与提问目的性.
Abstract
In order to improve the targeted selection of interactive students in the classroom questioning process,a recommendation method for individual classroom interactive students based on expression sequence vector clustering is proposed.Firstly,use an expression recognition model to obtain expression vectors from student side videos,summarize all expression vectors of each student to the teacher side,and combine them into an expression vector sequence;Then,based on the sequence of student expression vectors,a student classification method based on short-term expression vector clustering and a student recommendation method based on long-term expression vector clustering are proposed;Finally,presenting the classification and recommendation results of students to teachers through visualization and table format,providing sup-port for teachers to quickly select the most prominent individual student from different types.Through teacher evaluation,it was found that the proposed method can effectively express the differences in student states and significantly improve the discrimination efficiency and question-ing purpose of students in classroom interaction.
关键词
推荐方法/课堂交互/表情向量序列/数据可视化Key words
recommendation methods/classroom interaction/emotion vector sequence/data visualization引用本文复制引用
出版年
2024
国家科技期刊平台
NSTL
万方数据