@inproceedings{friidhriksdottir-2024-genderqueer,
title = "The {G}ender{Q}ueer Test Suite",
author = "Friidhriksd{\'o}ttir, Steinunn Rut",
editor = "Haddow, Barry and
Kocmi, Tom and
Koehn, Philipp and
Monz, Christof",
booktitle = "Proceedings of the Ninth Conference on Machine Translation",
month = nov,
year = "2024",
address = "Miami, Florida, USA",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.wmt-1.26",
pages = "327--340",
abstract = "This paper introduces the GenderQueer Test Suite, an evaluation set for assessing machine translation (MT) systems{'} capabilities in handling gender-diverse and queer-inclusive content, focusing on English to Icelandic translation. The suite evaluates MT systems on various aspects of gender-inclusive translation, including pronoun and adjective agreement, LGBTQIA+ terminology, and the impact of explicit gender specifications.The 17 MT systems submitted to the WMT24 English-Icelandic track were evaluated. Key findings reveal significant performance differences between large language model-based systems (LLMs) and lightweight models in handling context for gender agreement. Challenges in translating the singular {``}they{''} were widespread, while most systems performed relatively well in translating LGBTQIA+ terminology. Accuracy in adjective gender agreement is quite low, with some models struggling particularly with the feminine form.This evaluation set contributes to the ongoing discussion about inclusive language in MT and natural language processing. By providing a tool for assessing MT systems{'} handling of gender-diverse content, it aims to enhance the inclusivity of language technology. The methodology and evaluation scripts are made available for adaptation to other languages, promoting further research in this area.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="friidhriksdottir-2024-genderqueer">
<titleInfo>
<title>The GenderQueer Test Suite</title>
</titleInfo>
<name type="personal">
<namePart type="given">Steinunn</namePart>
<namePart type="given">Rut</namePart>
<namePart type="family">Friidhriksdóttir</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2024-11</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the Ninth Conference on Machine Translation</title>
</titleInfo>
<name type="personal">
<namePart type="given">Barry</namePart>
<namePart type="family">Haddow</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tom</namePart>
<namePart type="family">Kocmi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Philipp</namePart>
<namePart type="family">Koehn</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Christof</namePart>
<namePart type="family">Monz</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Miami, Florida, USA</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>This paper introduces the GenderQueer Test Suite, an evaluation set for assessing machine translation (MT) systems’ capabilities in handling gender-diverse and queer-inclusive content, focusing on English to Icelandic translation. The suite evaluates MT systems on various aspects of gender-inclusive translation, including pronoun and adjective agreement, LGBTQIA+ terminology, and the impact of explicit gender specifications.The 17 MT systems submitted to the WMT24 English-Icelandic track were evaluated. Key findings reveal significant performance differences between large language model-based systems (LLMs) and lightweight models in handling context for gender agreement. Challenges in translating the singular “they” were widespread, while most systems performed relatively well in translating LGBTQIA+ terminology. Accuracy in adjective gender agreement is quite low, with some models struggling particularly with the feminine form.This evaluation set contributes to the ongoing discussion about inclusive language in MT and natural language processing. By providing a tool for assessing MT systems’ handling of gender-diverse content, it aims to enhance the inclusivity of language technology. The methodology and evaluation scripts are made available for adaptation to other languages, promoting further research in this area.</abstract>
<identifier type="citekey">friidhriksdottir-2024-genderqueer</identifier>
<location>
<url>https://aclanthology.org/2024.wmt-1.26</url>
</location>
<part>
<date>2024-11</date>
<extent unit="page">
<start>327</start>
<end>340</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T The GenderQueer Test Suite
%A Friidhriksdóttir, Steinunn Rut
%Y Haddow, Barry
%Y Kocmi, Tom
%Y Koehn, Philipp
%Y Monz, Christof
%S Proceedings of the Ninth Conference on Machine Translation
%D 2024
%8 November
%I Association for Computational Linguistics
%C Miami, Florida, USA
%F friidhriksdottir-2024-genderqueer
%X This paper introduces the GenderQueer Test Suite, an evaluation set for assessing machine translation (MT) systems’ capabilities in handling gender-diverse and queer-inclusive content, focusing on English to Icelandic translation. The suite evaluates MT systems on various aspects of gender-inclusive translation, including pronoun and adjective agreement, LGBTQIA+ terminology, and the impact of explicit gender specifications.The 17 MT systems submitted to the WMT24 English-Icelandic track were evaluated. Key findings reveal significant performance differences between large language model-based systems (LLMs) and lightweight models in handling context for gender agreement. Challenges in translating the singular “they” were widespread, while most systems performed relatively well in translating LGBTQIA+ terminology. Accuracy in adjective gender agreement is quite low, with some models struggling particularly with the feminine form.This evaluation set contributes to the ongoing discussion about inclusive language in MT and natural language processing. By providing a tool for assessing MT systems’ handling of gender-diverse content, it aims to enhance the inclusivity of language technology. The methodology and evaluation scripts are made available for adaptation to other languages, promoting further research in this area.
%U https://aclanthology.org/2024.wmt-1.26
%P 327-340
Markdown (Informal)
[The GenderQueer Test Suite](https://aclanthology.org/2024.wmt-1.26) (Friidhriksdóttir, WMT 2024)
ACL
- Steinunn Rut Friidhriksdóttir. 2024. The GenderQueer Test Suite. In Proceedings of the Ninth Conference on Machine Translation, pages 327–340, Miami, Florida, USA. Association for Computational Linguistics.