信息服务

内容情感视角下的虚假短视频传播影响因素研究——基于CAC与ELM双模型

  • 付少雄 ,
  • 成琦
展开
  • (南京农业大学信息管理学院江苏210095)
付少雄,男,1993年生,南京农业大学信息管理学院副教授,硕士生导师。 成琦,男,2002年生,南京农业大学信息管理学院硕士研究生。

网络出版日期: 2025-03-19

基金资助

本文系国家社会科学基金青年项目“信息操纵影响下短视频虚假信息传播行为及其干预研究”(批准号:22CTQ018)的研究成果之一。

Research on the Influencing Factors of False Short Video Dissemination from the Perspective of Content Emotion: Based on CAC and ELM Dual Models

  • Fu Shaoxiong ,
  • Cheng Qi
Expand
  • (College of Information Management, Nanjing Agricultural University, Jiangsu, 210095)

Online published: 2025-03-19

摘要

[目的/意义]短视频已逐渐成为公众获取信息的主要方式之一。短视频平台中各类信息在用户间传播扩散,其中不乏大量虚假信息被用户采纳或分享,危害用户并扰乱网络信息环境。探索虚假短视频传播的影响因素,对虚假短视频治理和网络信息生态优化具有重要意义。[方法/过程]基于感知-情感-意向(Cognition-Affect-Conation,CAC)模型与精细加工可能性(Elaboration Likelihood Model,ELM)模型构建理论模型,以用户情绪为中介,以内容相关度为调节变量,通过抖音平台采集1007条有效虚假短视频数据,采用文本情感倾向分析、图像识别等技术提取虚假短视频的图像、音频与文本情感,探究虚假短视频内容情感对用户情绪及短视频传播效果的影响。[结果/结论]虚假短视频图像、音频与文本情感在不同程度上显著正向影响用户情绪,作者头像、背景音乐对用户情绪的影响不显著。用户情绪显著负向影响虚假短视频传播效果,在虚假短视频内容情感与其传播效果间具有部分中介作用。内容相关度显著调节用户情绪对虚假短视频传播效果的影响。

本文引用格式

付少雄 , 成琦 . 内容情感视角下的虚假短视频传播影响因素研究——基于CAC与ELM双模型[J]. 情报资料工作, 2025 , 46(2) : 61 -69 . DOI: 10.12154/j.qbzlgz.2025.02.007

Abstract

 [Purpose/significance] Short videos have gradually become one of the primary channels for the public to ob⁃tain information. Through short videos, various types of information can rapidly disseminate among users, including a large number of false short videos. Many of false short videos are adopted or shared by users, it may harm them and dis⁃turb the online information environment. Therefore, this study explores the factors influencing the dissemination of false short videos, in order to provide references for the governance of false short video and the optimization of online in⁃formation ecology. [Method/process] To explore the influence of content sentiment of false short video on user emotion and false short video dissemination effect, this study employed user sentiment as a mediator and content relevance as a moderator, constructed the research model based on the CAC Model and ELM Model, and then collected 1007 effective false short videos from the TikTok, extracted image, text and audio sentiment from these false short videos based on text
sentiment analysis, image recognition, and other technologies. As such, the study verified the moderating effect of con⁃tent relevance. [Result/conclusion] Images, text and audio sentiment of false short videos exert significant positive ef⁃fects on users' emotions in different degrees, while the author's avatar and background music exert no significant effect on users' emotions. User emotion exerts a significant negative effect on dissemination effect of false short videos, and partly plays a mediating role between the content sentiment of false short video and its dissemination effect. Content rel⁃evance significantly moderates the influence of the negative impact of user emotion on false short video dissemination effect.
文章导航

/