基于预训练融合模型的高校网络舆情分析研究

Research on College Network Public Opinion Analysis Based on Pre-training Fusion Model

针对预训练语言模型在情感分析领域中的应用进行分析,根据高校网络环境中舆情信息情感分类任务的实际需求,将传统的预训练语言模型进行融合,进一步提高模型分类预测的准确性。提出了一套针对高校网络舆情情感分析问题的系统设计方案,形成了从数据采集、数据处理到数据分析的完整闭环,能够有效辅助高校思政教育中对高校网络环境的舆情监督工作。

The application of pre-training language model in the field of sentiment analysis is analyzed,the traditional pre-training language model according to the actual needs of sentiment classification of public opinion information in the college network environment is integrated,to further improve the accuracy of model classification and prediction.A system design scheme for emotional analysis of the college network public opinion is put forward,which forms a complete closed-loop from data collection,data processing to data analysis and can effectively assist in public opinion supervision on the college network environment in ideological and political education.