图书情报工作 ›› 2021, Vol. 65 ›› Issue (23): 106-115.DOI: 10.13266/j.issn.0252-3116.2021.23.012

• 情报研究 • 上一篇    下一篇

网络舆情观点团簇演化等级测度与实证研究

闫璐, 杨刚, 赵江元   

  1. 吉林大学管理学院 长春 130022
  • 收稿日期:2021-06-20 修回日期:2021-09-23 出版日期:2021-12-05 发布日期:2021-12-18
  • 作者简介:闫璐,博士研究生,E-mail:623716346@qq.com;杨刚,教授,博士,博士生导师;赵江元,博士研究生。

Measurement and Empirical Study on the Evolution Level of Opinion Clusters of Internet Public Opinion

Yan Lu, Yang Gang, Zhao Jiangyuan   

  1. School of management, Jilin University, Changchun 130022
  • Received:2021-06-20 Revised:2021-09-23 Online:2021-12-05 Published:2021-12-18

摘要: [目的/意义]提出和构建网络舆情观点团簇演化等级,以描述网络舆情受众的群体性观点的状态随时间与事态变化的演化程度,对于网络舆情导控与精准引导具有重要的理论及实践意义。[方法/过程]基于LDA与CNN神经网络构建网络舆情观点团簇演化等级测度模型,并以"翟天临知网事件"为实验对象,验证演化等级这一指标的有效性。[结果/结论]网络舆情观点团簇演化等级能够很好地体现网络热点事件群体观点状态的演化,在展现3个维度的属性数值同时也能反映观点团簇较前一时间节点状态的演化程度,提出的观点团簇演化等级测度结果精准地体现事件观点的各个演化高峰,为有关部门对网络舆情群体观点的靶向引导提供新的指导方向。

关键词: 网络舆情, 观点团簇, 观点演化, 演化等级, 测度模型

Abstract: [Purpose/significance] Proposing and constructing the evolution level of Internet public opinion clusters is to describe the evolution degree of the group opinion state of Internet public opinion audiences over time and events. It is of great theoretical and practical significance for Internet public opinion guidance and precise guidance.[Method/process] Based on LDA and CNN neural Internet, the paper constructed a level measurement model of Internet public opinion cluster evolution, and took "Zhai Tianlin CNKI event" as the experimental object to verify the effectiveness of the index of evolution level.[Result/conclusion] The evolution level of Internet public opinion cluster can well reflect the evolution of Internet hot event group opinion state. It can show the attribute values of three dimensions and also reflect the evolution degree of opinion clusters compared with the node state of the previous time. The evolution level measurement results of opinion cluster in this paper accurately reflect each evolution peak of event opinion. It provides a new direction for the relevant departments to target and guide the opinions of Internet public opinion groups.

Key words: Internet public opinion, opinion cluster, opinion evolution, evolution level, measurement model

中图分类号: