数据与计算发展前沿 ›› 2024, Vol. 6 ›› Issue (6): 130-145.

CSTR: 32002.14.jfdc.CN10-1649/TP.2024.06.013

doi: 10.11871/jfdc.issn.2096-742X.2024.06.013

• • 上一篇    下一篇

社交媒体主题演化“热触发”效应与特征——基于新浪微博的数据挖掘

徐翔*(),徐舟爽,李嘉怡   

  1. 同济大学艺术与传媒学院，上海 201801
  • 收稿日期:2023-10-26 出版日期:2024-12-20 发布日期:2024-12-20
  • 通讯作者: 徐翔
  • 作者简介:徐翔，同济大学艺术与传媒学院教授、副院长，主要研究方向为社交媒体挖掘。
    本文主要承担工作为设计研究方法，制定论文框架，数据采集与文本挖掘，论文的修改与审定。
    XU Xiang is a full professor and the deputy Ddean of the School of Arts and Media, Tongji University. His main research interests include social media mining.
    In this paper, he is responsible for designing the research methods, formulating the paper framework, conducting data collection and text mining, and revising and finalizing the manuscript.|E-mail: xuxiang210089@163.com
  • 基金资助:
    国家自然科学基金(71804126);中央高校基本科研业务费专项资金资助(22120240589);同济大学实验教改专项(0600104109)

The “Heat trigger” Effect and Characteristics of Social Media Theme Evolution： Data Mining Based on Sina Weibo

XU Xiang*(),XU Zhoushuang,LI Jiayi   

  1. College of Arts and Media, Tongji University, Tongji University, Shanghai 201801, China
  • Received:2023-10-26 Online:2024-12-20 Published:2024-12-20
  • Contact: XU Xiang

RichHTML

4

PDF

320

摘要：

【目的】通过探讨社交媒体平台中主题热度的相互影响关系和演化机制，为平台信息窄化的形成路径与舆论集群调控提供新视角。热点“搭便车”“媒体风暴”等效应未关注到冷门主题之间的热度传递；现有研究多关注单一话题的内容、热度随时间演化规律，鲜有全局主题之间的影响网络研究；关于“热趋同”的研究忽略了相似度较低的主题间的热传递关系。【方法】以新浪微博1年中共20,117,408条帖子为样本，抽取不同主题热度在1年中的逐日变化序列，分析和检验微博中主题的平均热度对主题导致其他主题热度改变能力的影响。【结果】一个主题的热度变化会牵引其他的主题热度发生改变。该主题的平均热度越高，其热度变化带动、影响其他主题热度改变的能力就越强，也越容易被其他主题的热度改变影响自身热度。【结论】社交媒体平台中不同主题的热度之间存在规律性的相互影响关系。内容迥异的主题之间、冷门主题之间、冷门主题与热门主题之间的热度传导也是需要探讨的方面。

关键词: 社交媒体, 主题演化, 热触发, 平台信息茧房

Abstract:

[Objective] By exploring the mutual influence and evolution mechanism of topic popularity on social media platforms, this paper provides a new perspective on the formation path of platform information narrowing and the regulation of public opinion clusters. Hot topics “free riding”, “media storm” and other effects do not pay attention to the heat transfer between unpopular topics; Most of the existing studies focus on the content of a single topic and the evolution of its popularity over time, and few studies on the influence network between global topics. Studies on “thermal convergence” ignore the heat transfer relationship between subjects with low similarity. [Methods] A total of 20,117,408 posts on Sina Weibo in one year were taken as samples, and the daily change sequence of the popularity of different topics in one year was extracted to test the influence of the average popularity of topics in Weibo on the ability of topics to change the popularity of other topics. [Results] The change in the heat of one topic will bring about a change in the heat of other topics. The higher the average heat of the theme, the stronger the ability of its heat change to drive and affect the heat change of other topics, and the easier it is to be affected by the heat change of other topics. [Conclusions] There is a regular interaction between the popularity of different topics on social media platforms. The heat conduction between different topics, unpopular topics, and between unpopular topics and popular topics is also an aspect that needs to be discussed.

Key words: social media, theme evolution, heat trigger, information cocoons