|国家预印本平台
首页|GWO-LSTM预测下的高校网络舆情风险演化研究

GWO-LSTM预测下的高校网络舆情风险演化研究

Research on College online public opinion risk based improved GWO and LSTM

中文摘要英文摘要

研究目的 提高高校网络舆情事件风险预测的准确性对于维护国家安全和社会稳定具有重要意义。 研究方法 本研究基于微博平台中高校网络舆情事件的热度和评论,提出了一种基于舆情事件热度和情感分析结果的舆情风险模型,使用灰狼算法与LSTM相结合作为预测模型,分析高校网络舆情风险演化的情况,并结合案例进行了分析验证。 研究结论 本文构建了全链条高校网络舆情分析框架,模型考虑因素更为全面,预测精度和拟合度高,能全面地反映实际舆情的风险态势及其变化,防控演化分析表明利用模型预测舆情可及时将舆情事件控制,防止持续发酵。研究表明,该模型能很好地反映并预测舆情风险的程度与变化,在验证事件中,平均绝对误差为13.8%,同时利用模型的演化分析论证了高校网络舆情预测的重要性和必要性。

Research objective Improving the accuracy of online public opinion event risk prediction in universities is of great significance for maintaining national security and social stability. Research method Based on the popularity and comments of online public opinion events in colleges and universities on Weibo platform, this study proposed a public opinion risk model based on the popularity of public opinion events and sentiment analysis results. The Grey Wolf algorithm combined with LSTM was used as a prediction model to analyze the evolution of online public opinion risks in colleges and universities, and analyzed and verified with cases. Research Conclusion This paper builds a whole-chain online public opinion analysis framework in universities. The model has more comprehensive considerations, high prediction accuracy and fitting degree, and can fully reflect the risk situation and changes of actual public opinion. The analysis of prevention and control evolution shows that using the model to predict public opinion can control public opinion events in time and prevent continuous fermentation. The research shows that the model can well reflect and predict the degree and change of public opinion risk, and the average absolute error is 13.8% in the verification event. At the same time, the importance and necessity of online public opinion prediction in universities are demonstrated by the evolutionary analysis of the model.

10.12074/202405.00032V1

信息传播、知识传播教育计算技术、计算机技术

高校网络舆情舆情风险LSTM情感分析灰狼算法舆情演化

University public opinion onlinePublic opinion riskLSTMEmotion analysisGrey Wolf algorithmEvolution of public opinion

.GWO-LSTM预测下的高校网络舆情风险演化研究[EB/OL].(2024-05-06)[2025-07-21].https://chinaxiv.org/abs/202405.00032.点此复制

评论