摘要
在数字化学习环境中,拥有具身形象的教育智能体可以承担人类教师的部分社会功能,从而改善学习者社会交互不足的问题。不过,已有研究发现,教育智能体的积极情绪线索并不总是能唤起学习者的积极情绪,其原因可能与情绪的测量方法有关。具体而言,人类的情绪具有波动性和衰弱性,量表等事后测量工具难以准确地记录学习者的即时情绪反应。为有效测量教育智能体情绪线索对学习者的影响,设计采用主观(量表)、客观(面部表情识别)相结合的方式,旨在测量高校学习者在不同学习情境中的过程性情绪反应、结果性情绪感知以及他们的学习动机。实验数据分析结果表明:积极的情绪线索可以显著提升学习者的内在目标倾向和任务价值;积极的情绪线索可以促进学习者的积极情绪,并且基于面部表情的测量方法,可以更加精准地反馈学习者的过程性情绪,尤其是对教育智能体情绪线索的即时反应。该研究结果在实践层面,扩充了对教育智能体情绪线索作用效果的理解;在方法层面,应用了智能技术支撑的及时性、精细化测量方法,为智能化的情绪感知、计算与调节奠定了基础。
In digitized learning environment,pedagogical agent with embodiment can partially take the social functions of human instructors,and address learners’lack of social interactions.However,previous studies found that positive emotional cues of pedagogical agents did not always induce learners’positive emotions,which might be related to how emotions were measured.Specifically,human emotions may change and calm down rather quickly,thus post-hoc measurements such as survey cannot accurately record learners’emotional reactions.To further discover the effects of pedagogical agent emotional cues on learners,62 college students from multiple majors were recurited in an experiment.Subjective(survey)and objective(facial expression recognition)were employed to measure learners’perceived emotions and motivations after learning,as well as their immediate emotional reactions during learning.Data analysis results indicated that(1)positive emotional cues can significantly increase learners’intrinsic goal orientation and task value;(2)positive emotional cues can promote learners’positive emotions,and facial expression-based measurement can more accurately reflect learners’immediate emotions,especially their emotional reactions to pedagogical agent’s emotional cues.This research,at the practical level,broadens current understandings towards pedagogical agent emotional cues,while at the methodological level,it has applied method of AI-supported immediate and refined measurement,which laid the foundation for intelligent emotion perception,calculation,and adjustment.
作者
巴深
刘清堂
吴林静
于钦春
Ba Shen;Liu Qingtang;Wu Linjing;Yu Qinchun(Faculty of Artificial Intelligence in Education,Central China Normal University,Wuhan Hubei 430079)
出处
《远程教育杂志》
CSSCI
北大核心
2021年第6期48-57,共10页
Journal of Distance Education
基金
2017年度国家自然科学基金项目“非数学语言描述问题的机器理解方法研究”(项目编号:61772012)
湖北省教育科学规划2020年度重点课题“构建线上线下融合的课堂新型教学模式创新实证研究”(项目编号:2020GA005)研究成果。
关键词
教育智能体
智能教育
情绪线索
多模态学习
开放学习资源
Pedagogical Agent
Intelligent Education
Emotional Cues
Multimodal Learning
Open Learning Resources
