摘要
[目的/意义]高校网络舆情能够反映学生的思想动态。实时甄别高校网络信息内容,监测潜在舆情,对于合理引导学生思想、维护高校形象具有重要意义。[方法/过程]提出基于LDA-BiLSTM模型的高校网络舆情监测方法,以高校百度贴吧为数据源,提取数据中的热门主题,获取热门主题下回复文本的情感极性,分析其潜在舆情。[结果/结论]结果表明,此方法能够获取高校贴吧热门主题及其情感极性,监测潜在舆情,为高校网络舆情管理提供有效方法和模型。
[Purpose/significance]Network public opinion in colleges and universities can reflect students’ideological trends.It is of great significance to properly guide students’thoughts and maintain the image of colleges and universities to screen the network information contents of colleges and to monitor potential public opinions in real time.[Method/process]A monitoring method of network public opinion in colleges and universities based on LDA-BiLSTM model was proposed.With Baidu Tieba in colleges and universities as the data source.Extract the hot topics of data and get the emotional polarity of the reply texts under hot topics.Analyze potential public opinions.[Result/conclusion]The results show that the method can obtain hot topics and their emotional polarity of Tieba in colleges and universities,monitor potential public opinions.To provide effective methods and models for network public opinion management in colleges and universities.
出处
《情报理论与实践》
CSSCI
北大核心
2020年第11期156-161,共6页
Information Studies:Theory & Application
基金
国家自然科学基金青年项目“考虑风险依赖和竞争外部性的信息安全外包决策机制和策略优化研究”(项目编号:71801125)
2019年校党建与思政教育研究课题二类资助课题“高校网络舆情事件应对研究”(项目编号:SZ20190202)的成果。
关键词
高校学生
网络舆情
隐含狄里克雷分布
长短时记忆网络
舆情监测
college students
network public opinion
Latent Dirichlet Allocation
long short term memory
public opinion monitoring
