A Study of Misinformation in Audio Messages Shared in WhatsApp Groups

被引:6
作者
Maros, Alexandre [1 ]
Almeida, Jussara M. [1 ]
Vasconcelos, Marisa [2 ]
机构
[1] Univ Fed Minas Gerais, Belo Horizonte, MG, Brazil
[2] IBM Res, Sao Paulo, Brazil
来源
DISINFORMATION IN OPEN ONLINE MEDIA, MISDOOM 2021 | 2021年 / 12887卷
关键词
WhatsApp; Audio messages; Misinformation;
D O I
10.1007/978-3-030-87031-7_6
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Recent studies have shown that group communication on WhatsApp plays a significant role to foster information dissemination at large, with evidence of its use for misinformation campaigns. We analyze more than 40K audio messages shared in over 364 publicly accessible groups in Brazil, covering six months of great social mobilization in the country. We identify the presence of misinformation in these audios by relying on previously checked facts. Our study focuses on content and propagation properties of audio misinformation, contrasting them with unchecked content as well as with prior findings of misinformation in other media types. We also rely on a set of volunteers to perform a qualitative analysis of the audios. We observed that audios with misinformation had a higher presence of negative emotions and also often used phrases in the future tense and talked directly to the listener. Moreover, audios with misinformation tend to spread quicker than unchecked content and last significantly longer in the network. The speaker's tone from the audios with misinformation was also considered less friendly and natural than the unchecked ones. Our study contributes to the literature by focusing on a media type that is gaining mainstream popularity recently, and, as we show here, is being used as vessel for misinformation spread.
引用
收藏
页码:85 / 100
页数:16
相关论文
共 23 条
[1]   Social Media and Fake News in the 2016 Election [J].
Allcott, Hunt ;
Gentzkow, Matthew .
JOURNAL OF ECONOMIC PERSPECTIVES, 2017, 31 (02) :211-235
[2]  
[Anonymous], 2013, Int. J. Adv. Res. Comput. Sci. Softw. Eng.
[3]   Detecting misinformation in social networks using provenance data [J].
Baeth, Mohamed Jehad ;
Aktas, Mehmet S. .
CONCURRENCY AND COMPUTATION-PRACTICE & EXPERIENCE, 2019, 31 (03)
[4]  
Bessi A., 2016, 1 MONDAY, V21, P14, DOI DOI 10.5210/FM.V21I11.7090
[5]   Latent Dirichlet allocation [J].
Blei, DM ;
Ng, AY ;
Jordan, MI .
JOURNAL OF MACHINE LEARNING RESEARCH, 2003, 3 (4-5) :993-1022
[6]   Weaponized Health Communication: Twitter Bots and Russian Trolls Amplify the Vaccine Debate [J].
Broniatowski, David A. ;
Jamison, Amelia M. ;
Qi, SiHua ;
AlKulaib, Lulwah ;
Chen, Tao ;
Benton, Adrian ;
Quinn, Sandra C. ;
Dredze, Mark .
AMERICAN JOURNAL OF PUBLIC HEALTH, 2018, 108 (10) :1378-1384
[7]   Thousands of Small, Constant Rallies: A Large-Scale Analysis of Partisan WhatsApp Groups [J].
Bursztyn, Victor S. ;
Birnbaum, Larry .
PROCEEDINGS OF THE 2019 IEEE/ACM INTERNATIONAL CONFERENCE ON ADVANCES IN SOCIAL NETWORKS ANALYSIS AND MINING (ASONAM 2019), 2019, :484-488
[8]   Characterizing Attention Cascades in WhatsApp Groups [J].
Caetano, Josemar Alves ;
Magno, Gabriel ;
Goncalves, Marcos ;
Almeida, Jussara ;
Marques-Neto, Humberto T. ;
Almeida, Virgilio .
PROCEEDINGS OF THE 11TH ACM CONFERENCE ON WEB SCIENCE (WEBSCI'19), 2019, :27-36
[9]   Supervised machine learning for audio emotion recognition Enhancing film sound design using audio features, regression models and artificial neural networks [J].
Cunningham, Stuart ;
Ridley, Harrison ;
Weinel, Jonathan ;
Picking, Richard .
PERSONAL AND UBIQUITOUS COMPUTING, 2021, 25 (04) :637-650
[10]  
Fernando Gavin, 2018, NEWS COM AU 0719