【作 者】：
马玉慧， 夏雪莹， 张文慧
【关 键 词 】：
课堂提问； 自动分类； 深度神经网络； 深度学习
【栏 目】：
学科建设与教师发展
【中文摘要】：
课堂提问是教师课堂教学行为的关键组成部分，是师生进行课堂交互的主要方式。对教师课堂提问进行分析，是提升教师课堂教学水平的关键。视频分析法是目前进行课堂提问分析的主要方法。但该方法需要花费大量的时间和人力，导致无法进行大规模应用。近几年，随着人工智能技术的不断成熟，越来越多的领域开始利用人工智能替代人工操作。为使课堂提问分析能够高效、大规模地应用，本研究提出基于深度学习的课堂提问自动分析方法。研究采用卷积神经网络（CNN）及长短时记忆网络（LSTM），对80节课的9090条课堂教师提问文本进行分类。实验结果表明，CNN模型具有更好的分类效果，其在提问内容和提问类型两个维度上的整体准确率分别是85.17%和87.84%。应用该方法训练的模型，可替代传统的视频分析法，能够实现大规模的教师课堂提问话语的自动分析。
【英文摘要】：
Classroom questioning is a key component of teachers' classroom teaching behavior and the main way for teachers and students to interact in class. To analyze teachers' classroom questions is the key to improve teachers' classroom teaching. Currently, video analysis is the main method used to analyze classroom questions. But this method takes a lot of time and manpower, which makes it impossible to apply on a large scale. In recent years, with the continuous maturity of artificial intelligence technology, artificial intelligence is used in more and more fields to replace manual operation. In order to analyze classroom questions efficiently and on a large scale, this study proposes an automatic analysis method based on deep learning. The experimental results show that the CNN model has a better classification effect, with an overall accuracy of 85.17% and 87.84% in two dimensions of question content and question type respectively. The model trained by this method can replace the traditional video analysis method and realize large-scale automatic analysis of teachers' classroom questioning discourse.