@inproceedings{parrish-etal-2024-diversity,
title = "Diversity-Aware Annotation for Conversational {AI} Safety",
author = "Parrish, Alicia and
Prabhakaran, Vinodkumar and
Aroyo, Lora and
D{\'\i}az, Mark and
Homan, Christopher M. and
Serapio-Garc{\'\i}a, Greg and
Taylor, Alex S. and
Wang, Ding",
editor = "Dinkar, Tanvi and
Attanasio, Giuseppe and
Curry, Amanda Cercas and
Konstas, Ioannis and
Hovy, Dirk and
Rieser, Verena",
booktitle = "Proceedings of Safety4ConvAI: The Third Workshop on Safety for Conversational AI @ LREC-COLING 2024",
month = may,
year = "2024",
address = "Torino, Italia",
publisher = "ELRA and ICCL",
url = "https://aclanthology.org/2024.safety4convai-1.2",
pages = "8--15",
abstract = "How people interpret content is deeply influenced by their socio-cultural backgrounds and lived experiences. This is especially crucial when evaluating AI systems for safety, where accounting for such diversity in interpretations and potential impacts on human users will make them both more successful and inclusive. While recent work has demonstrated the importance of diversity in human ratings that underlie AI pipelines, effective and efficient ways to incorporate diverse perspectives in human data annotation pipelines is still largely elusive. In this paper, we discuss the primary challenges faced in incorporating diversity into model evaluations, and propose a practical diversity-aware annotation approach. Using an existing dataset with highly parallel safety annotations, we take as a test case a policy that prioritizes recall of safety issues, and demonstrate that our diversity-aware approach can efficiently obtain a higher recall of safety issues flagged by minoritized rater groups without hurting overall precision.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="parrish-etal-2024-diversity">
<titleInfo>
<title>Diversity-Aware Annotation for Conversational AI Safety</title>
</titleInfo>
<name type="personal">
<namePart type="given">Alicia</namePart>
<namePart type="family">Parrish</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Vinodkumar</namePart>
<namePart type="family">Prabhakaran</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Lora</namePart>
<namePart type="family">Aroyo</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mark</namePart>
<namePart type="family">Díaz</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Christopher</namePart>
<namePart type="given">M</namePart>
<namePart type="family">Homan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Greg</namePart>
<namePart type="family">Serapio-García</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Alex</namePart>
<namePart type="given">S</namePart>
<namePart type="family">Taylor</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ding</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2024-05</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of Safety4ConvAI: The Third Workshop on Safety for Conversational AI @ LREC-COLING 2024</title>
</titleInfo>
<name type="personal">
<namePart type="given">Tanvi</namePart>
<namePart type="family">Dinkar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Giuseppe</namePart>
<namePart type="family">Attanasio</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Amanda</namePart>
<namePart type="given">Cercas</namePart>
<namePart type="family">Curry</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ioannis</namePart>
<namePart type="family">Konstas</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Dirk</namePart>
<namePart type="family">Hovy</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Verena</namePart>
<namePart type="family">Rieser</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>ELRA and ICCL</publisher>
<place>
<placeTerm type="text">Torino, Italia</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>How people interpret content is deeply influenced by their socio-cultural backgrounds and lived experiences. This is especially crucial when evaluating AI systems for safety, where accounting for such diversity in interpretations and potential impacts on human users will make them both more successful and inclusive. While recent work has demonstrated the importance of diversity in human ratings that underlie AI pipelines, effective and efficient ways to incorporate diverse perspectives in human data annotation pipelines is still largely elusive. In this paper, we discuss the primary challenges faced in incorporating diversity into model evaluations, and propose a practical diversity-aware annotation approach. Using an existing dataset with highly parallel safety annotations, we take as a test case a policy that prioritizes recall of safety issues, and demonstrate that our diversity-aware approach can efficiently obtain a higher recall of safety issues flagged by minoritized rater groups without hurting overall precision.</abstract>
<identifier type="citekey">parrish-etal-2024-diversity</identifier>
<location>
<url>https://aclanthology.org/2024.safety4convai-1.2</url>
</location>
<part>
<date>2024-05</date>
<extent unit="page">
<start>8</start>
<end>15</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Diversity-Aware Annotation for Conversational AI Safety
%A Parrish, Alicia
%A Prabhakaran, Vinodkumar
%A Aroyo, Lora
%A Díaz, Mark
%A Homan, Christopher M.
%A Serapio-García, Greg
%A Taylor, Alex S.
%A Wang, Ding
%Y Dinkar, Tanvi
%Y Attanasio, Giuseppe
%Y Curry, Amanda Cercas
%Y Konstas, Ioannis
%Y Hovy, Dirk
%Y Rieser, Verena
%S Proceedings of Safety4ConvAI: The Third Workshop on Safety for Conversational AI @ LREC-COLING 2024
%D 2024
%8 May
%I ELRA and ICCL
%C Torino, Italia
%F parrish-etal-2024-diversity
%X How people interpret content is deeply influenced by their socio-cultural backgrounds and lived experiences. This is especially crucial when evaluating AI systems for safety, where accounting for such diversity in interpretations and potential impacts on human users will make them both more successful and inclusive. While recent work has demonstrated the importance of diversity in human ratings that underlie AI pipelines, effective and efficient ways to incorporate diverse perspectives in human data annotation pipelines is still largely elusive. In this paper, we discuss the primary challenges faced in incorporating diversity into model evaluations, and propose a practical diversity-aware annotation approach. Using an existing dataset with highly parallel safety annotations, we take as a test case a policy that prioritizes recall of safety issues, and demonstrate that our diversity-aware approach can efficiently obtain a higher recall of safety issues flagged by minoritized rater groups without hurting overall precision.
%U https://aclanthology.org/2024.safety4convai-1.2
%P 8-15
Markdown (Informal)
[Diversity-Aware Annotation for Conversational AI Safety](https://aclanthology.org/2024.safety4convai-1.2) (Parrish et al., Safety4ConvAI-WS 2024)
ACL
- Alicia Parrish, Vinodkumar Prabhakaran, Lora Aroyo, Mark Díaz, Christopher M. Homan, Greg Serapio-García, Alex S. Taylor, and Ding Wang. 2024. Diversity-Aware Annotation for Conversational AI Safety. In Proceedings of Safety4ConvAI: The Third Workshop on Safety for Conversational AI @ LREC-COLING 2024, pages 8–15, Torino, Italia. ELRA and ICCL.