图书情报工作 ›› 2019, Vol. 63 ›› Issue (14): 62-73.DOI: 10.13266/j.issn.0252-3116.2019.14.008

• 情报研究 • 上一篇    下一篇

社交网络舆情意见领袖研究:蝴蝶图示、甄别及影响力评价

郭勇1, 高歌1,2,3, 王天勇4, Harris Wu5, 周世玉6   

  1. 1. 吉林大学管理学院 长春 130022;
    2. 维也纳经济大学生产运作系 维也纳 1100;
    3. 吉林大学珠海学院 珠海 519000;
    4. 广东珠海横琴新区建设环保局 珠海 519000;
    5. 美国欧道明大学商学院 诺福克 23008;
    6. 台湾师范大学管理学院 台北 10610
  • 收稿日期:2019-01-13 修回日期:2019-04-15 出版日期:2019-07-20 发布日期:2019-07-20
  • 通讯作者: 高歌(ORCID:0000-0002-5881-319X),教授,硕士生导师,维也纳经济大学研究员,通讯作者,E-mail:gaogew@126.com
  • 作者简介:郭勇(ORCID:0000-0001-7888-3294),本科生;王天勇(ORCID:0000-0003-1658-8761),副教授;Harris Wu (ORCID:0000-0002-8686-8585),教授,博士生导师;周世玉(ORCID:0000-0003-0956-6222),教授,博士生导师。
  • 基金资助:
    本文系国家自然科学基金面上项目"大数据环境下多媒体网络舆情信息的语义识别与危机响应研究"(项目编号:71473101)研究成果之一。

Research on Social Network Opinion Leaders: Butterfly Schema, Identification and Impact Assessment

Guo Yong1, Gao Ge1,2,3, Wang Tianyong4, Harris Wu5, Shiyu Chou6   

  1. 1. School of Management, Jilin University, Changchun 130022;
    2. Department of Production and Operation, Vienna University of Economics and Business, Vienna 1100;
    3. Zhuhai College, Jilin University, Zhuhai 519000;
    4. Hengqin Construction Environmental Protection Bureau, Zhuhai 519000;
    5. Strome College of Business, Old Dominion University, Norfolk 23008;
    6. College of Management, Taiwan Normal University, Taipei 10610
  • Received:2019-01-13 Revised:2019-04-15 Online:2019-07-20 Published:2019-07-20

摘要: [目的/意义]社交媒体环境下意见领袖与受众社群间形成了自运转、自循环的范围舆情系统,有效甄别意见领袖并评价其影响力对加强网络舆情管控具有实践意义。[方法/过程]结合OCA扩展理论、群际关系理论、舆情场势理论以及SIC理论,通过系统动力学分析意见领袖作用、前因变量及其动因机制,构建意见领袖影响力评价体系的一般性框架,提出一种变权重灰色关联度的意见领袖甄别算法,并以舆情话题"11·3留日女生遇害案"进行实证研究。[结果/结论]"蝴蝶图示"架构了意见领袖作用与前因变量的因果关系及反馈回路,揭示了舆情系统内社群生态与意见领袖作用协同演化的内在机理;本文提出的理论模型具有多维度测度、权重集科学、算法性能优越等特点,适用于社交网络中意见领袖形成的动态过程。

关键词: 意见领袖, 灰色关联度, 指标体系, 网络舆情

Abstract: [Purpose/significance] The self-operation and circulation system of public opinion has come into being between opinion leaders and audience communities in social media environment. It's beneficial for the government's supervision of network public opinion to study the identification of opinion leaders and the evaluation of their influence.[Method/process] Combining with the theories of OCA extension, intergroup relations, public opinion field potential and the social identity construction, this paper analyzes the roles of opinion leaders, antecedent variables and efficient cause mechanism with the system dynamic methods. In addition, this paper constructs a general framework for the evaluation system of opinion leaders' impact. Meanwhile, it also puts forward the leader opinion identification algorithm based on the variable weight grey correlation degree. Finally, an empirical study is proved with the case of "11.3 Murder of Girl Studying in Japan".[Result/conclusion] "Butterfly Schema" builds the causal relationship and feedback loop between opinion leaders and antecedents and reveals the inherent mechanism of the co-evolution of community ecology and opinion leaders in public opinion system. The theoretical model proposed in this paper has the characteristics of multi-dimensional measurement, scientific weight set and superior algorithm performance, which is suitable for the dynamic process of opinion leaders formation in social networks.

Key words: opinion leaders, grey correlation degree, index system, network public opinion

中图分类号: