基于LDA-BiLSTM模型的高校网络舆情监测方法及实证

Method and Demonstration of University Network Public Opinion Monitoring Based on LDA-BiLSTM Model

[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.