图书情报工作 ›› 2021, Vol. 65 ›› Issue (9): 70-78.DOI: 10.13266/j.issn.0252-3116.2021.09.008

• 情报研究 • 上一篇    下一篇

社交媒体虚假健康信息特征识别

张帅   

  1. 武汉大学信息管理学院 武汉 430072 武汉大学信息资源研究中心 武汉 430072
  • 收稿日期:2020-11-15 修回日期:2021-02-02 出版日期:2021-05-05 发布日期:2021-06-02
  • 作者简介:张帅(ORCID:0000-0002-5792-877X),博士研究生,E-mail:zs09053@163.com。
  • 基金资助:
    本文系湖北文化名家专项基金资助研究成果之一。

Study on Feature Identification of False Health Information on Social Media

Zhang Shuai   

  1. School of Information Management, Wuhan University, Wuhan 430072 Center for Studies of Information Resources, Wuhan University, Wuhan 430072
  • Received:2020-11-15 Revised:2021-02-02 Online:2021-05-05 Published:2021-06-02

摘要: [目的/意义] 识别社交媒体虚假健康信息特征,构建社交媒体虚假健康信息特征清单,以期为社交媒体虚假健康信息特征的测度提供一定理论支撑,也为用户和社交媒体平台判别虚假健康信息提供有益参考。[方法/过程] 采集1 004条社交媒体健康数据,利用程序化编码抽取社交媒体虚假健康信息的关键特征,运用卡方检验和方差分析揭示社交媒体虚假健康信息的显著特征,并构建社交媒体虚假健康信息特征清单。[结果/结论] 研究结果表明,社交媒体虚假健康信息特征具有表面特征、语义特征和来源特征3个维度、11个主要特征以及29个子特征。其中,社交媒体上食品安全主题的虚假健康信息在"术语包装"特征上表现得更为显著;"夸大事实"为社交媒体上常见疾病主题虚假健康信息的显著特征;社交媒体上养生保健主题的虚假健康信息具有"元数据缺失"和"假借权威"显著特征。

关键词: 社交媒体, 健康信息, 虚假特征, 特征识别

Abstract: [Purpose/significance] This study identifies the features of false health information on social media and construct a list of false health information characteristics on social media, in order to provide certain theoretical support for the measurement of false health information features on social media, and also provide a useful reference for users and social media service platforms to distinguish false health information.[Method/process] 1 004 pieces of empirical data from social media were collected, and the key features of false health information were extracted by programmatic coding. Then the chi-square test and analysis of variance were adopted to identify significant features of health misinformation. In addition, this study developed a list of features to identify health misinformation on social media.[Result/conclusion] It was shown that the features of false health information on social media had three dimensions:surface features, semantic features, and source features. There were 11 main features and 29 sub-features. It was found that the features of "term packaging" on food safety topic was more notable than other topics; "exaggerated facts" on common diseases topic was more significant than other topics; the features of "lack of meta-information" and "fake authority" on healthcare topic were more prominent than other topics.

Key words: social media, health information, false features, feature identification

中图分类号: