Thyghgapp: Instagram content moderation and lexical variation in Pro-Eating disorder communities

Stevie Chancellor; Jessica Pater; Trustin Clear; Eric Gilbert; Munmun De Choudhury

doi:10.1145/2818048.2819963

Thyghgapp: Instagram content moderation and lexical variation in Pro-Eating disorder communities

Stevie Chancellor, Jessica Pater, Trustin Clear, Eric Gilbert, Munmun De Choudhury

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

206 Scopus citations

Abstract

Pro-eating disorder (pro-ED) communities on social media encourage the adoption and maintenance of disordered eating habits as acceptable alternative lifestyles rather than threats to health. In particular, the social networking site Instagram has reacted by banning searches on several pro- ED tags and issuing content advisories on others. We present the first large-scale quantitative study investigating pro-ED communities on Instagram in the aftermath of moderation - our dataset contains 2.5M posts between 2011 and 2014. We find that the pro-ED community has adopted nonstandard lexical variations of moderated tags to circumvent these restrictions. In fact, increasingly complex lexical variants have emerged over time. Communities that use lexical variants show increased participation and support of pro- ED (15-30%). Finally, the tags associated with content on these variants express more toxic, self-harm, and vulnerable content. Despite Instagram's moderation strategies, pro-ED communities are active and thriving. We discuss the effectiveness of content moderation as an intervention for communities of deviant behavior.

Original language	English (US)
Title of host publication	Proceedings of the 19th ACM Conference on Computer-Supported Cooperative Work and Social Computing, CSCW 2016
Publisher	Association for Computing Machinery
Pages	1201-1213
Number of pages	13
ISBN (Electronic)	9781450335928
DOIs	https://doi.org/10.1145/2818048.2819963
State	Published - Feb 27 2016
Externally published	Yes
Event	19th ACM Conference on Computer-Supported Cooperative Work and Social Computing, CSCW 2016 - San Francisco, United States Duration: Feb 27 2016 → Mar 2 2016

Publication series

Name	Proceedings of the ACM Conference on Computer Supported Cooperative Work, CSCW
Volume	27

Other

Other	19th ACM Conference on Computer-Supported Cooperative Work and Social Computing, CSCW 2016
Country/Territory	United States
City	San Francisco
Period	2/27/16 → 3/2/16

Bibliographical note

Publisher Copyright:
© 2016 ACM.

Keywords

Eating disorder
Instagram
Lexical variation
Social media

Publisher link

10.1145/2818048.2819963

Find It

Find It at U of M Libraries

Cite this

Chancellor, S., Pater, J., Clear, T., Gilbert, E., & De Choudhury, M. (2016). Thyghgapp: Instagram content moderation and lexical variation in Pro-Eating disorder communities. In Proceedings of the 19th ACM Conference on Computer-Supported Cooperative Work and Social Computing, CSCW 2016 (pp. 1201-1213). (Proceedings of the ACM Conference on Computer Supported Cooperative Work, CSCW; Vol. 27). Association for Computing Machinery. https://doi.org/10.1145/2818048.2819963

Thyghgapp: Instagram content moderation and lexical variation in Pro-Eating disorder communities. / Chancellor, Stevie; Pater, Jessica; Clear, Trustin et al.
Proceedings of the 19th ACM Conference on Computer-Supported Cooperative Work and Social Computing, CSCW 2016. Association for Computing Machinery, 2016. p. 1201-1213 (Proceedings of the ACM Conference on Computer Supported Cooperative Work, CSCW; Vol. 27).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Chancellor, S, Pater, J, Clear, T, Gilbert, E & De Choudhury, M 2016, Thyghgapp: Instagram content moderation and lexical variation in Pro-Eating disorder communities. in Proceedings of the 19th ACM Conference on Computer-Supported Cooperative Work and Social Computing, CSCW 2016. Proceedings of the ACM Conference on Computer Supported Cooperative Work, CSCW, vol. 27, Association for Computing Machinery, pp. 1201-1213, 19th ACM Conference on Computer-Supported Cooperative Work and Social Computing, CSCW 2016, San Francisco, United States, 2/27/16. https://doi.org/10.1145/2818048.2819963

Chancellor S, Pater J, Clear T, Gilbert E, De Choudhury M. Thyghgapp: Instagram content moderation and lexical variation in Pro-Eating disorder communities. In Proceedings of the 19th ACM Conference on Computer-Supported Cooperative Work and Social Computing, CSCW 2016. Association for Computing Machinery. 2016. p. 1201-1213. (Proceedings of the ACM Conference on Computer Supported Cooperative Work, CSCW). doi: 10.1145/2818048.2819963

Chancellor, Stevie ; Pater, Jessica ; Clear, Trustin et al. / Thyghgapp : Instagram content moderation and lexical variation in Pro-Eating disorder communities. Proceedings of the 19th ACM Conference on Computer-Supported Cooperative Work and Social Computing, CSCW 2016. Association for Computing Machinery, 2016. pp. 1201-1213 (Proceedings of the ACM Conference on Computer Supported Cooperative Work, CSCW).

@inproceedings{1f8da77d7d484f43b559304e29a9e91d,

title = "Thyghgapp: Instagram content moderation and lexical variation in Pro-Eating disorder communities",

abstract = "Pro-eating disorder (pro-ED) communities on social media encourage the adoption and maintenance of disordered eating habits as acceptable alternative lifestyles rather than threats to health. In particular, the social networking site Instagram has reacted by banning searches on several pro- ED tags and issuing content advisories on others. We present the first large-scale quantitative study investigating pro-ED communities on Instagram in the aftermath of moderation - our dataset contains 2.5M posts between 2011 and 2014. We find that the pro-ED community has adopted nonstandard lexical variations of moderated tags to circumvent these restrictions. In fact, increasingly complex lexical variants have emerged over time. Communities that use lexical variants show increased participation and support of pro- ED (15-30%). Finally, the tags associated with content on these variants express more toxic, self-harm, and vulnerable content. Despite Instagram's moderation strategies, pro-ED communities are active and thriving. We discuss the effectiveness of content moderation as an intervention for communities of deviant behavior.",

keywords = "Eating disorder, Instagram, Lexical variation, Social media",

author = "Stevie Chancellor and Jessica Pater and Trustin Clear and Eric Gilbert and {De Choudhury}, Munmun",

note = "Publisher Copyright: {\textcopyright} 2016 ACM.; 19th ACM Conference on Computer-Supported Cooperative Work and Social Computing, CSCW 2016 ; Conference date: 27-02-2016 Through 02-03-2016",

year = "2016",

month = feb,

day = "27",

doi = "10.1145/2818048.2819963",

language = "English (US)",

series = "Proceedings of the ACM Conference on Computer Supported Cooperative Work, CSCW",

publisher = "Association for Computing Machinery",

pages = "1201--1213",

booktitle = "Proceedings of the 19th ACM Conference on Computer-Supported Cooperative Work and Social Computing, CSCW 2016",

}

TY - GEN

T1 - Thyghgapp

T2 - 19th ACM Conference on Computer-Supported Cooperative Work and Social Computing, CSCW 2016

AU - Chancellor, Stevie

AU - Pater, Jessica

AU - Clear, Trustin

AU - Gilbert, Eric

AU - De Choudhury, Munmun

PY - 2016/2/27

Y1 - 2016/2/27

N2 - Pro-eating disorder (pro-ED) communities on social media encourage the adoption and maintenance of disordered eating habits as acceptable alternative lifestyles rather than threats to health. In particular, the social networking site Instagram has reacted by banning searches on several pro- ED tags and issuing content advisories on others. We present the first large-scale quantitative study investigating pro-ED communities on Instagram in the aftermath of moderation - our dataset contains 2.5M posts between 2011 and 2014. We find that the pro-ED community has adopted nonstandard lexical variations of moderated tags to circumvent these restrictions. In fact, increasingly complex lexical variants have emerged over time. Communities that use lexical variants show increased participation and support of pro- ED (15-30%). Finally, the tags associated with content on these variants express more toxic, self-harm, and vulnerable content. Despite Instagram's moderation strategies, pro-ED communities are active and thriving. We discuss the effectiveness of content moderation as an intervention for communities of deviant behavior.

AB - Pro-eating disorder (pro-ED) communities on social media encourage the adoption and maintenance of disordered eating habits as acceptable alternative lifestyles rather than threats to health. In particular, the social networking site Instagram has reacted by banning searches on several pro- ED tags and issuing content advisories on others. We present the first large-scale quantitative study investigating pro-ED communities on Instagram in the aftermath of moderation - our dataset contains 2.5M posts between 2011 and 2014. We find that the pro-ED community has adopted nonstandard lexical variations of moderated tags to circumvent these restrictions. In fact, increasingly complex lexical variants have emerged over time. Communities that use lexical variants show increased participation and support of pro- ED (15-30%). Finally, the tags associated with content on these variants express more toxic, self-harm, and vulnerable content. Despite Instagram's moderation strategies, pro-ED communities are active and thriving. We discuss the effectiveness of content moderation as an intervention for communities of deviant behavior.

KW - Eating disorder

KW - Instagram

KW - Lexical variation

KW - Social media

UR - http://www.scopus.com/inward/record.url?scp=84963576979&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84963576979&partnerID=8YFLogxK

U2 - 10.1145/2818048.2819963

DO - 10.1145/2818048.2819963

M3 - Conference contribution

AN - SCOPUS:84963576979

T3 - Proceedings of the ACM Conference on Computer Supported Cooperative Work, CSCW

SP - 1201

EP - 1213

BT - Proceedings of the 19th ACM Conference on Computer-Supported Cooperative Work and Social Computing, CSCW 2016

PB - Association for Computing Machinery

Y2 - 27 February 2016 through 2 March 2016

ER -

Thyghgapp: Instagram content moderation and lexical variation in Pro-Eating disorder communities

Abstract

Publication series

Other

Bibliographical note

Keywords

Publisher link

Find It

Other files and links

Fingerprint

Cite this