安徽省高校就业社会网络舆情预测模型研究
Study on Forecasting the Social Network Public Opinion of College Employment in Anhui Province
云坡 1刘程慧 1方小枝1
作者信息
- 1. 合肥大学 经济与管理学院,安徽 合肥 230601
- 折叠
摘要
预测就业网络舆情有助于跟踪掌握就业舆情变动,助力政府就业政策的精准出台.使用百度指数,构造安徽省高等院校就业社会网络舆情信息的文本词频集,构建机器学习EEMD-GRU混合模型,对就业社会网络舆情进行拟合与预测.结果显示:EEMD-GRU模型能有效刻画安徽省高校就业社会网络舆情趋势,揭示舆情信息的多尺度时频特征,预测误差RMSE、MAE、MAPE仅为 0.971、0.773、0.229,呈现较高准确度.这表明模型能为政府部门研判高校就业舆情、制定政策提供量化分析支撑.
Abstract
Studying and predicting the social network public opinion on college graduates'employment will help to track and grasp the trend of public opinions on employment,and help the government to introduce precise employment policies.By using Baidu index to construct the text word frequency collection of the public opinion information of the employment social network of colleges and universities in Anhui Province,and building a machine learning EEMD-GRU hybrid model,nonlinear fitting and mapping of social attention of employment rate are carried out.The results show that the EEMD-GRU model can describe and reveal the multi-scale time-frequency characteristics of the social network public opinion of college employment in Anhui province,and the prediction errors of RMSE,MAE and MAPE are only 0.971,0.773 and 0.229,showing high prediction accuracy and stability,indicating that the model can provide a technical research and judgment basis for public opinion of college employment and the formulation of government employment policies.
关键词
社会网络舆情/EEMD-GRU/预测Key words
social network public opinion/EEMD-GRU/predicting引用本文复制引用
基金项目
教育部人文社会科学研究青年基金(21YJC790152)
安徽省省级质量工程"会计学专业改造提升"项目(2021zygzts054)
合肥大学质量工程教研项目(2022hfujyzd09)
出版年
2024
NETL
NSTL
万方数据