@inproceedings{kivlichan-etal-2021-measuring,
title = "Measuring and Improving Model-Moderator Collaboration using Uncertainty Estimation",
author = "Kivlichan, Ian and
Lin, Zi and
Liu, Jeremiah and
Vasserman, Lucy",
editor = "Mostafazadeh Davani, Aida and
Kiela, Douwe and
Lambert, Mathias and
Vidgen, Bertie and
Prabhakaran, Vinodkumar and
Waseem, Zeerak",
booktitle = "Proceedings of the 5th Workshop on Online Abuse and Harms (WOAH 2021)",
month = aug,
year = "2021",
address = "Online",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2021.woah-1.5",
doi = "10.18653/v1/2021.woah-1.5",
pages = "36--53",
abstract = "Content moderation is often performed by a collaboration between humans and machine learning models. However, it is not well understood how to design the collaborative process so as to maximize the combined moderator-model system performance. This work presents a rigorous study of this problem, focusing on an approach that incorporates model uncertainty into the collaborative process. First, we introduce principled metrics to describe the performance of the collaborative system under capacity constraints on the human moderator, quantifying how efficiently the combined system utilizes human decisions. Using these metrics, we conduct a large benchmark study evaluating the performance of state-of-the-art uncertainty models under different collaborative review strategies. We find that an uncertainty-based strategy consistently outperforms the widely used strategy based on toxicity scores, and moreover that the choice of review strategy drastically changes the overall system performance. Our results demonstrate the importance of rigorous metrics for understanding and developing effective moderator-model systems for content moderation, as well as the utility of uncertainty estimation in this domain.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="kivlichan-etal-2021-measuring">
<titleInfo>
<title>Measuring and Improving Model-Moderator Collaboration using Uncertainty Estimation</title>
</titleInfo>
<name type="personal">
<namePart type="given">Ian</namePart>
<namePart type="family">Kivlichan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zi</namePart>
<namePart type="family">Lin</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jeremiah</namePart>
<namePart type="family">Liu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Lucy</namePart>
<namePart type="family">Vasserman</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2021-08</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 5th Workshop on Online Abuse and Harms (WOAH 2021)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Aida</namePart>
<namePart type="family">Mostafazadeh Davani</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Douwe</namePart>
<namePart type="family">Kiela</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mathias</namePart>
<namePart type="family">Lambert</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Bertie</namePart>
<namePart type="family">Vidgen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Vinodkumar</namePart>
<namePart type="family">Prabhakaran</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zeerak</namePart>
<namePart type="family">Waseem</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Online</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Content moderation is often performed by a collaboration between humans and machine learning models. However, it is not well understood how to design the collaborative process so as to maximize the combined moderator-model system performance. This work presents a rigorous study of this problem, focusing on an approach that incorporates model uncertainty into the collaborative process. First, we introduce principled metrics to describe the performance of the collaborative system under capacity constraints on the human moderator, quantifying how efficiently the combined system utilizes human decisions. Using these metrics, we conduct a large benchmark study evaluating the performance of state-of-the-art uncertainty models under different collaborative review strategies. We find that an uncertainty-based strategy consistently outperforms the widely used strategy based on toxicity scores, and moreover that the choice of review strategy drastically changes the overall system performance. Our results demonstrate the importance of rigorous metrics for understanding and developing effective moderator-model systems for content moderation, as well as the utility of uncertainty estimation in this domain.</abstract>
<identifier type="citekey">kivlichan-etal-2021-measuring</identifier>
<identifier type="doi">10.18653/v1/2021.woah-1.5</identifier>
<location>
<url>https://aclanthology.org/2021.woah-1.5</url>
</location>
<part>
<date>2021-08</date>
<extent unit="page">
<start>36</start>
<end>53</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Measuring and Improving Model-Moderator Collaboration using Uncertainty Estimation
%A Kivlichan, Ian
%A Lin, Zi
%A Liu, Jeremiah
%A Vasserman, Lucy
%Y Mostafazadeh Davani, Aida
%Y Kiela, Douwe
%Y Lambert, Mathias
%Y Vidgen, Bertie
%Y Prabhakaran, Vinodkumar
%Y Waseem, Zeerak
%S Proceedings of the 5th Workshop on Online Abuse and Harms (WOAH 2021)
%D 2021
%8 August
%I Association for Computational Linguistics
%C Online
%F kivlichan-etal-2021-measuring
%X Content moderation is often performed by a collaboration between humans and machine learning models. However, it is not well understood how to design the collaborative process so as to maximize the combined moderator-model system performance. This work presents a rigorous study of this problem, focusing on an approach that incorporates model uncertainty into the collaborative process. First, we introduce principled metrics to describe the performance of the collaborative system under capacity constraints on the human moderator, quantifying how efficiently the combined system utilizes human decisions. Using these metrics, we conduct a large benchmark study evaluating the performance of state-of-the-art uncertainty models under different collaborative review strategies. We find that an uncertainty-based strategy consistently outperforms the widely used strategy based on toxicity scores, and moreover that the choice of review strategy drastically changes the overall system performance. Our results demonstrate the importance of rigorous metrics for understanding and developing effective moderator-model systems for content moderation, as well as the utility of uncertainty estimation in this domain.
%R 10.18653/v1/2021.woah-1.5
%U https://aclanthology.org/2021.woah-1.5
%U https://doi.org/10.18653/v1/2021.woah-1.5
%P 36-53
Markdown (Informal)
[Measuring and Improving Model-Moderator Collaboration using Uncertainty Estimation](https://aclanthology.org/2021.woah-1.5) (Kivlichan et al., WOAH 2021)
ACL