Explainability and hate speech: Structured explanations make social media moderators faster

Agostina Calabrese; Leonardo Neves; Neil Shah; Maarten W. Bos; Björn Ross; Mirella Lapata; Francesco Barbieri

doi:10.48550/arXiv.2406.04106

Explainability and hate speech: Structured explanations make social media moderators faster

Agostina Calabrese^*, Leonardo Neves, Neil Shah, Maarten W. Bos, Björn Ross, Mirella Lapata, Francesco Barbieri

^*Corresponding author for this work

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

Abstract

Content moderators play a key role in keeping the conversation on social media healthy. While the high volume of content they need to judge represents a bottleneck to the moderation pipeline, no studies have explored how models could support them to make faster decisions. There is, by now, a vast body of research into detecting hate speech, sometimes explicitly motivated by a desire to help improve content moderation, but published research using real content moderators is scarce. In this work we investigate the effect of explanations on the speed of real-world moderators. Our experiments show that while generic explanations do not affect their speed and are often ignored, structured explanations lower moderators' decision making time by 7.4%.

Original language	English
Title of host publication	Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics
Publisher	ACL Anthology
DOIs	https://doi.org/10.48550/arXiv.2406.04106
Publication status	Accepted/In press - 16 May 2024
Event	The 62nd Annual Meeting of the Association for Computational Linguistics - Centara Grand and Bangkok Convention Centre at CentralWorld, Bangkok, Thailand Duration: 11 Aug 2024 → 16 Aug 2024 Conference number: 62 https://2024.aclweb.org/

Conference

Conference	The 62nd Annual Meeting of the Association for Computational Linguistics
Abbreviated title	ACL 2024
Country/Territory	Thailand
City	Bangkok
Period	11/08/24 → 16/08/24
Internet address	https://2024.aclweb.org/

Access to Document

10.48550/arXiv.2406.04106Licence: Creative Commons: Attribution (CC-BY)

CalabreseEtalACL2024ExplainabilityAndHateSpeechAccepted author manuscript, 1.32 MBLicence: Creative Commons: Attribution (CC-BY)

1 Finished
1 Active

TEAMER : Teaching Machines to Reason Like Humans
Lapata, M. (Principal Investigator)
UK central government bodies/local authorities, health and hospital authorities
1/10/21 → 30/09/26
Project: Research
TransModal: Translating from Multiple Modalities into Text
Lapata, M. (Principal Investigator)
EU government bodies
1/09/16 → 31/08/22
Project: Research

Cite this

@inproceedings{d74d797c82444f34b6f1167f32055a56,

title = "Explainability and hate speech: Structured explanations make social media moderators faster",

abstract = "Content moderators play a key role in keeping the conversation on social media healthy. While the high volume of content they need to judge represents a bottleneck to the moderation pipeline, no studies have explored how models could support them to make faster decisions. There is, by now, a vast body of research into detecting hate speech, sometimes explicitly motivated by a desire to help improve content moderation, but published research using real content moderators is scarce. In this work we investigate the effect of explanations on the speed of real-world moderators. Our experiments show that while generic explanations do not affect their speed and are often ignored, structured explanations lower moderators' decision making time by 7.4%.",

author = "Agostina Calabrese and Leonardo Neves and Neil Shah and Bos, {Maarten W.} and Bj{\"o}rn Ross and Mirella Lapata and Francesco Barbieri",

year = "2024",

month = may,

day = "16",

doi = "10.48550/arXiv.2406.04106",

language = "English",

booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics",

publisher = "ACL Anthology",

note = "The 62nd Annual Meeting of the Association for Computational Linguistics, ACL 2024 ; Conference date: 11-08-2024 Through 16-08-2024",

url = "https://2024.aclweb.org/",

}

Calabrese, A, Neves, L, Shah, N, Bos, MW, Ross, B , Lapata, M & Barbieri, F 2024, Explainability and hate speech: Structured explanations make social media moderators faster. in Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics. ACL Anthology, The 62nd Annual Meeting of the Association for Computational Linguistics, Bangkok, Thailand, 11/08/24. https://doi.org/10.48550/arXiv.2406.04106

TY - GEN

T1 - Explainability and hate speech

T2 - The 62nd Annual Meeting of the Association for Computational Linguistics

AU - Calabrese, Agostina

AU - Neves, Leonardo

AU - Shah, Neil

AU - Bos, Maarten W.

AU - Ross, Björn

AU - Lapata, Mirella

AU - Barbieri, Francesco

N1 - Conference code: 62

PY - 2024/5/16

Y1 - 2024/5/16

N2 - Content moderators play a key role in keeping the conversation on social media healthy. While the high volume of content they need to judge represents a bottleneck to the moderation pipeline, no studies have explored how models could support them to make faster decisions. There is, by now, a vast body of research into detecting hate speech, sometimes explicitly motivated by a desire to help improve content moderation, but published research using real content moderators is scarce. In this work we investigate the effect of explanations on the speed of real-world moderators. Our experiments show that while generic explanations do not affect their speed and are often ignored, structured explanations lower moderators' decision making time by 7.4%.

AB - Content moderators play a key role in keeping the conversation on social media healthy. While the high volume of content they need to judge represents a bottleneck to the moderation pipeline, no studies have explored how models could support them to make faster decisions. There is, by now, a vast body of research into detecting hate speech, sometimes explicitly motivated by a desire to help improve content moderation, but published research using real content moderators is scarce. In this work we investigate the effect of explanations on the speed of real-world moderators. Our experiments show that while generic explanations do not affect their speed and are often ignored, structured explanations lower moderators' decision making time by 7.4%.

U2 - 10.48550/arXiv.2406.04106

DO - 10.48550/arXiv.2406.04106

M3 - Conference contribution

BT - Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics

PB - ACL Anthology

Y2 - 11 August 2024 through 16 August 2024

ER -

Explainability and hate speech: Structured explanations make social media moderators faster

Abstract

Conference

Access to Document

Fingerprint

Projects

TEAMER : Teaching Machines to Reason Like Humans

TransModal: Translating from Multiple Modalities into Text

Cite this