佛山科学技术学院学报（自然科学版）2025，Vol.43Issue(3)：61-68,8.
基于LogSparse Transformer模型的高校网络舆情预测与分析
University network public opinion prediction and analysis based on LogSparse Transformer model
摘要
Abstract
A prediction method based on LogSparse Transformer time series model is proposed to address the rapid changes and complexity of online public opinion in universities,in order to improve the accuracy and efficiency of public opinion management.The study used data from the Sina Weibo platform and constructed a LogSparse Transformer model through data preprocessing,correlation mining of long-term data,and long-range dependency modeling.The experimental results show that the LogSparse Transformer model outperforms traditional methods and machine learning algorithms in terms of prediction accuracy,while also having faster response speed and real-time processing capabilities.This model can effectively capture long-range dependencies in university public opinion events and reduce the time complexity of the model.This model provides a new and effective tool for predicting and managing public opinion in universities,which helps university managers to respond and manage online public opinion in a timely manner.关键词
网络舆情/LogSparse Transformer时序模型/高校管理/数据清洗/注意力机制Key words
online public opinion/LogSparse Transformer time series model/university management/data cleaning/attention mechanism分类
计算机与自动化引用本文复制引用
张友海..基于LogSparse Transformer模型的高校网络舆情预测与分析[J].佛山科学技术学院学报（自然科学版）,2025,43(3):61-68,8.基金项目
安徽省高校科学研究重点项目(2022AH052053) （2022AH052053）
