情感分析在学生评教中的应用

Application of Sentiment Analysis in Student Assessment

摘要: 为适应教育信息化的发展趋势，各高校将部分传统教学环节转移到线上，因此每年会产生大量的教学管理及评学评教的数据，其中，分析学生评教数据是教师了解学生需求，提高教学质量的有效途径。论文利用某高校真实的学生评教数据，探索学生对各教学环节的关注与情感倾向。论文首先对学生评教的历史与现状进行说明；再结合SnowNLP情感分析的特点，利用情感词典生成情感分析模型；将数据预处理后，通过模型获得评教数据的正负性；再通过统计分析的方法，将结果可视化呈现。分析发现，学生对备课最关注且积极评价比例最高，而对作业评讲环节关注度最低，课堂管理与课堂互动环节消极评价比例最高。分析结果可为教师改善教学过程提供参考依据。

Abstract: In order to meet the development trend of education informatization, universities have transferred some traditional teaching sessions online, so a large amount of data on teaching management and evaluation of learning and teaching will be generated every year, among which student evaluation data is an effective way for teachers to understand students’ needs and improve teaching quality. The paper uses real student evaluation data from a university to explore students' concerns and emotional tendencies towards each teaching session. The paper first explains the history and current situation of students’ evaluation of teaching; then combines the features of SnowNLP sentiment analysis and uses sentiment dictionary to generate sentiment analysis model; after pre-pro- cessing the data, the positive and negative aspects of the evaluation data are obtained through the model; and then the results are presented visually through statistical analysis. The analysis found that students were most concerned about lesson preparation and had the highest proportion of positive evaluations, while they were least concerned about the homework assessment session and had the highest proportion of negative evaluations in classroom management and classroom interaction. The analysis results can provide a reference basis for teachers to improve the teaching process.