【作 者】：
陈靓影， 罗珍珍， 徐如意
【关 键 词 】：
课堂环境； 学习兴趣智能分析； 多模态融合； 自然感知技术； 学习兴趣模型
【栏 目】：
课程与教学
【中文摘要】：
学习兴趣影响学习发生的方式和学习过程,是提高学习效果的重要因素。目前对教学环境下学生学习兴趣的研究大多采用传统的问卷调查或案例分析等方法，不利于教师对课堂中学生学习兴趣的及时获取以便有效改善教学行为。因此，本文基于多模态自然感知技术展开对学习兴趣的智能化分析研究，提出了三维学习兴趣模型，通过对头部姿态、面部表情、课堂互动多模态信息识别与融合，从认知注意力、学习情感与思维活跃度三个维度综合理解学生的学习兴趣。实验结果表明，该学习兴趣智能化分析方法能真实反映学生的学习兴趣状态，为教师改进教学方法、提高教学效果提供客观依据及有效手段。
【英文摘要】：
Learning interest, which affects the way and process of learning, is an important factor to improve the learning effect. At present, the study on students' learning interest in class mostly adopts traditional methods such as questionnaire survey, case analysis, which are not conductive for teachers to know students' interest timely and improve their teaching behaviors effectively. Therefore, this paper conducts an intelligent analysis of learning interest based on multi-modal natural perception technology, and proposes a three-dimensional learning interest model. Owing to this model, through recognition and fusion of multi-modal information of head posture, facial expression and classroom interaction, students' learning interests can be comprehensively understood from three dimensions of cognitive attention, learning emotion and thinking activity. The experimental results show that the intelligent analysis of learning interest can be used to reflect students' learning interest, which provides objective basis and effective means for teachers to improve their teaching methods and teaching effect.