@inproceedings{uz-etal-2026-turcomet,
title = "{T}ur{COMET}: A Morphologically-Aware Evaluation System for Document-Level {T}urkish-{E}nglish Machine Translation",
author = {Uz, Atahan and
Akda{\u{g}}, H{\"u}seyin Emir and
{\"U}sk{\"u}darl{\i}, Susan},
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.emnlp-main.2672/",
pages = "60806--60826",
ISBN = "979-8-89176-459-0",
abstract = "We introduce TurCOMET, the first Turkish-specific augmentation of COMET for evaluating translations between Turkish and English. By extending xCOMET-XL with Turkish morphological features and training it on additional Turkish-English data via an additive bottleneck layer, TurCOMET resolves neural evaluation failures caused by sub-word over-segmentation. To validate our metric, we introduce TurkDoc-MT, a meticulously curated Turkish-English document-level aligned dataset of 1,322 instances spanning seven domains. Meta-evaluation demonstrates that TurCOMET significantly outperforms the off-the-shelf baseline, yielding absolute Kendall{'}s {\ensuremath{\tau}} correlation gains of +0.113 (system-level) and +0.024 (segment-level), with the largest gains in the TR{\textrightarrow}EN direction. All models, datasets, code, and annotations are publicly released for the benefit of Turkish natural language processing and benchmarking any translation system."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="uz-etal-2026-turcomet">
<titleInfo>
<title>TurCOMET: A Morphologically-Aware Evaluation System for Document-Level Turkish-English Machine Translation</title>
</titleInfo>
<name type="personal">
<namePart type="given">Atahan</namePart>
<namePart type="family">Uz</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hüseyin</namePart>
<namePart type="given">Emir</namePart>
<namePart type="family">Akdağ</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Susan</namePart>
<namePart type="family">Üsküdarlı</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-10</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hung-Yi</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sunipa</namePart>
<namePart type="family">Dev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jessy</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Desmond</namePart>
<namePart type="family">Elliott</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Budapest, Hungary</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-459-0</identifier>
</relatedItem>
<abstract>We introduce TurCOMET, the first Turkish-specific augmentation of COMET for evaluating translations between Turkish and English. By extending xCOMET-XL with Turkish morphological features and training it on additional Turkish-English data via an additive bottleneck layer, TurCOMET resolves neural evaluation failures caused by sub-word over-segmentation. To validate our metric, we introduce TurkDoc-MT, a meticulously curated Turkish-English document-level aligned dataset of 1,322 instances spanning seven domains. Meta-evaluation demonstrates that TurCOMET significantly outperforms the off-the-shelf baseline, yielding absolute Kendall’s \ensuremathτ correlation gains of +0.113 (system-level) and +0.024 (segment-level), with the largest gains in the TR→EN direction. All models, datasets, code, and annotations are publicly released for the benefit of Turkish natural language processing and benchmarking any translation system.</abstract>
<identifier type="citekey">uz-etal-2026-turcomet</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-main.2672/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>60806</start>
<end>60826</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T TurCOMET: A Morphologically-Aware Evaluation System for Document-Level Turkish-English Machine Translation
%A Uz, Atahan
%A Akdağ, Hüseyin Emir
%A Üsküdarlı, Susan
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-459-0
%F uz-etal-2026-turcomet
%X We introduce TurCOMET, the first Turkish-specific augmentation of COMET for evaluating translations between Turkish and English. By extending xCOMET-XL with Turkish morphological features and training it on additional Turkish-English data via an additive bottleneck layer, TurCOMET resolves neural evaluation failures caused by sub-word over-segmentation. To validate our metric, we introduce TurkDoc-MT, a meticulously curated Turkish-English document-level aligned dataset of 1,322 instances spanning seven domains. Meta-evaluation demonstrates that TurCOMET significantly outperforms the off-the-shelf baseline, yielding absolute Kendall’s \ensuremathτ correlation gains of +0.113 (system-level) and +0.024 (segment-level), with the largest gains in the TR→EN direction. All models, datasets, code, and annotations are publicly released for the benefit of Turkish natural language processing and benchmarking any translation system.
%U https://aclanthology.org/2026.emnlp-main.2672/
%P 60806-60826
Markdown (Informal)
[TurCOMET: A Morphologically-Aware Evaluation System for Document-Level Turkish-English Machine Translation](https://aclanthology.org/2026.emnlp-main.2672/) (Uz et al., EMNLP 2026)
ACL