文献类型：期刊文献

英文题名：Sentiment analysis of course evaluation data based on SVM Model

作者：Zhao, Junyan[1]; Yang, Xubiao[2]; Qiao, Qian[3]; Chen, Liqiong[4]

机构：[1] East China University of Science and Technology, School of Network Education, Shanghai, China; [2] Shanghai University of Engineering Science, College of Air Transport, Shanghai, China; [3] East China University of Science and Technology, School of Social and Public Administration, Shanghai, China; [4] Shanghai Institute of Technology, Department of Computer Science and Information Engineering, Shanghai, China

年份：2020

起止页码：375

外文期刊名：Proceedings of 2020 IEEE International Conference on Progress in Informatics and Computing, PIC 2020

收录：EI(收录号：20210910005858)

语种：英文

外文关键词：Support vector machines - Character recognition - Classification (of information) - Function evaluation - Learning systems - Social aspects - Curricula

摘要：Machine learning is a hot technology in the field of modern computers and greatly affects people's way of life. As an important application of machine learning, emotion analysis can analyze people's emotional preferences and opinions on specific subjects, such as stock trends, public opinion analysis. In the traditional teaching evaluation system, the evaluation text cannot be directly quantified into scores, and the manual recognition needs a lot of effort, so the actual value is not big. This system uses crawler to crawl the evaluation text, and through the data enhancement, data cleaning, emotion dictionary construction, word vector conversion and other operations to complete the data preprocessing, and finally through the SVM model training and DAG multi-classification structure to complete the text emotion of five classification functions. On the basis of the algorithm, using the Flask to build network project achieve the function of frontend data display and interface design, thus helping the school to understand the students satisfaction with the degree of course teachers. ? 2020 IEEE.