摘要
针对传统企业网络舆情危机预警方法存在预警准确性较低的问题,提出基于迁移学习的企业网络舆情危机预警方法。利用智能元搜索技术对企业网络舆情信息进行分类,通过迁移学习分析网络舆情负面信息数据,提取网络舆情危机特征,识别网络舆情危机,并确定预警等级,实现危机预警。实验结果表明,该方法的预警性能最好,优于对照组。
Aiming at the problem of low accuracy in traditional enterprise network public opinion crisis warning methods,a transfer learning based enterprise network public opinion crisis warning method is proposed.Using intelligent meta search technology to classify enterprise network public opinion information,analyzing negative information data of network public opinion through transfer learning,extracting crisis characteristics of network public opinion,identifying network public opinion crises,and determining warning levels to achieve crisis warning.The experimental results show that the warning performance of this method is the best,better than the control group.
作者
卢贤玲
王梦露
孙滨
LU Xianling;WANG Menglu;SUN Bin(College of Information Engineering,Zhengzhou University of Industrial Technology,Xinzheng Henan 451100,China;Institute of Marxism,Zhengzhou University of Industrial Technology,Xinzheng Henan 451100,China)
出处
《信息与电脑》
2023年第21期38-40,共3页
Information & Computer
基金
河南省科技厅科技攻关支持项目(项目编号:222102210159,222102210224)
河南省科技厅软科学支持项目(项目编号:222400410228)
河南省哲学社会科学规划支持项目(项目编号:2022BXW012)。
关键词
迁移学习
企业网络舆情
危机预警
transfer learning
corporate online public opinion
crisis warning
