@inproceedings{eltahir-etal-2025-autoarabic,
title = "{A}uto{A}rabic: A Three-Stage Framework for Localizing Video-Text Retrieval Benchmarks",
author = "Eltahir, Mohamed and
Sarraj, Osamah and
Alfrihidi, Abdulrahman M. and
Alshatiri, Taha and
Khurd, Mohammed and
Bremoo, Mohammed and
Hussain, Tanveer",
editor = "Darwish, Kareem and
Ali, Ahmed and
Abu Farha, Ibrahim and
Touileb, Samia and
Zitouni, Imed and
Abdelali, Ahmed and
Al-Ghamdi, Sharefah and
Alkhereyf, Sakhar and
Zaghouani, Wajdi and
Khalifa, Salam and
AlKhamissi, Badr and
Almatham, Rawan and
Hamed, Injy and
Alyafeai, Zaid and
Alowisheq, Areeb and
Inoue, Go and
Mrini, Khalil and
Alshammari, Waad",
booktitle = "Proceedings of The Third Arabic Natural Language Processing Conference",
month = nov,
year = "2025",
address = "Suzhou, China",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.arabicnlp-main.23/",
pages = "288--297",
ISBN = "979-8-89176-352-4",
abstract = "Video-to-text and text-to-video retrieval are dominated by English benchmarks (e.g. DiDeMo, MSR-VTT) and recent multilingual corpora (e.g. RUDDER), yet Arabic remains underserved, lacking localized evaluation metrics. We introduce a three-stage framework, AutoArabic, utilizing state-of-the-art large language models (LLMs) to translate non-Arabic benchmarks into Modern Standard Arabic, reducing the manual revision required by nearly fourfold. The framework incorporates an error detection module that automatically flags potential translation errors with 97{\%} accuracy. Applying the framework to DiDeMo, a video retrieval benchmark produces DiDeMo-AR, an Arabic variant with 40,144 fluent Arabic descriptions. An analysis of the translation errors is provided and organized into an insightful taxonomy to guide future Arabic localization efforts. We train a CLIP-style baseline with identical hyperparameters on the Arabic and English variants of the benchmark, finding a moderate performance gap ($\Delta \approx 3$pp at Recall@1), indicating that Arabic localization preserves benchmark difficulty. We evaluate three post-editing budgets (zero/ flagged-only/ full) and find that performance improves monotonically with more post-editing, while the raw LLM output (zero-budget) remains usable. To ensure reproducibility to other languages, we made the code available at https://github.com/Tahaalshatiri/AutoArabic."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="eltahir-etal-2025-autoarabic">
<titleInfo>
<title>AutoArabic: A Three-Stage Framework for Localizing Video-Text Retrieval Benchmarks</title>
</titleInfo>
<name type="personal">
<namePart type="given">Mohamed</namePart>
<namePart type="family">Eltahir</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Osamah</namePart>
<namePart type="family">Sarraj</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Abdulrahman</namePart>
<namePart type="given">M</namePart>
<namePart type="family">Alfrihidi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Taha</namePart>
<namePart type="family">Alshatiri</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mohammed</namePart>
<namePart type="family">Khurd</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mohammed</namePart>
<namePart type="family">Bremoo</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tanveer</namePart>
<namePart type="family">Hussain</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-11</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of The Third Arabic Natural Language Processing Conference</title>
</titleInfo>
<name type="personal">
<namePart type="given">Kareem</namePart>
<namePart type="family">Darwish</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ahmed</namePart>
<namePart type="family">Ali</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ibrahim</namePart>
<namePart type="family">Abu Farha</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Samia</namePart>
<namePart type="family">Touileb</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Imed</namePart>
<namePart type="family">Zitouni</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ahmed</namePart>
<namePart type="family">Abdelali</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sharefah</namePart>
<namePart type="family">Al-Ghamdi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sakhar</namePart>
<namePart type="family">Alkhereyf</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Wajdi</namePart>
<namePart type="family">Zaghouani</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Salam</namePart>
<namePart type="family">Khalifa</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Badr</namePart>
<namePart type="family">AlKhamissi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Rawan</namePart>
<namePart type="family">Almatham</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Injy</namePart>
<namePart type="family">Hamed</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zaid</namePart>
<namePart type="family">Alyafeai</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Areeb</namePart>
<namePart type="family">Alowisheq</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Go</namePart>
<namePart type="family">Inoue</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Khalil</namePart>
<namePart type="family">Mrini</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Waad</namePart>
<namePart type="family">Alshammari</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Suzhou, China</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-352-4</identifier>
</relatedItem>
<abstract>Video-to-text and text-to-video retrieval are dominated by English benchmarks (e.g. DiDeMo, MSR-VTT) and recent multilingual corpora (e.g. RUDDER), yet Arabic remains underserved, lacking localized evaluation metrics. We introduce a three-stage framework, AutoArabic, utilizing state-of-the-art large language models (LLMs) to translate non-Arabic benchmarks into Modern Standard Arabic, reducing the manual revision required by nearly fourfold. The framework incorporates an error detection module that automatically flags potential translation errors with 97% accuracy. Applying the framework to DiDeMo, a video retrieval benchmark produces DiDeMo-AR, an Arabic variant with 40,144 fluent Arabic descriptions. An analysis of the translation errors is provided and organized into an insightful taxonomy to guide future Arabic localization efforts. We train a CLIP-style baseline with identical hyperparameters on the Arabic and English variants of the benchmark, finding a moderate performance gap (Δ \approx 3pp at Recall@1), indicating that Arabic localization preserves benchmark difficulty. We evaluate three post-editing budgets (zero/ flagged-only/ full) and find that performance improves monotonically with more post-editing, while the raw LLM output (zero-budget) remains usable. To ensure reproducibility to other languages, we made the code available at https://github.com/Tahaalshatiri/AutoArabic.</abstract>
<identifier type="citekey">eltahir-etal-2025-autoarabic</identifier>
<location>
<url>https://aclanthology.org/2025.arabicnlp-main.23/</url>
</location>
<part>
<date>2025-11</date>
<extent unit="page">
<start>288</start>
<end>297</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T AutoArabic: A Three-Stage Framework for Localizing Video-Text Retrieval Benchmarks
%A Eltahir, Mohamed
%A Sarraj, Osamah
%A Alfrihidi, Abdulrahman M.
%A Alshatiri, Taha
%A Khurd, Mohammed
%A Bremoo, Mohammed
%A Hussain, Tanveer
%Y Darwish, Kareem
%Y Ali, Ahmed
%Y Abu Farha, Ibrahim
%Y Touileb, Samia
%Y Zitouni, Imed
%Y Abdelali, Ahmed
%Y Al-Ghamdi, Sharefah
%Y Alkhereyf, Sakhar
%Y Zaghouani, Wajdi
%Y Khalifa, Salam
%Y AlKhamissi, Badr
%Y Almatham, Rawan
%Y Hamed, Injy
%Y Alyafeai, Zaid
%Y Alowisheq, Areeb
%Y Inoue, Go
%Y Mrini, Khalil
%Y Alshammari, Waad
%S Proceedings of The Third Arabic Natural Language Processing Conference
%D 2025
%8 November
%I Association for Computational Linguistics
%C Suzhou, China
%@ 979-8-89176-352-4
%F eltahir-etal-2025-autoarabic
%X Video-to-text and text-to-video retrieval are dominated by English benchmarks (e.g. DiDeMo, MSR-VTT) and recent multilingual corpora (e.g. RUDDER), yet Arabic remains underserved, lacking localized evaluation metrics. We introduce a three-stage framework, AutoArabic, utilizing state-of-the-art large language models (LLMs) to translate non-Arabic benchmarks into Modern Standard Arabic, reducing the manual revision required by nearly fourfold. The framework incorporates an error detection module that automatically flags potential translation errors with 97% accuracy. Applying the framework to DiDeMo, a video retrieval benchmark produces DiDeMo-AR, an Arabic variant with 40,144 fluent Arabic descriptions. An analysis of the translation errors is provided and organized into an insightful taxonomy to guide future Arabic localization efforts. We train a CLIP-style baseline with identical hyperparameters on the Arabic and English variants of the benchmark, finding a moderate performance gap (Δ \approx 3pp at Recall@1), indicating that Arabic localization preserves benchmark difficulty. We evaluate three post-editing budgets (zero/ flagged-only/ full) and find that performance improves monotonically with more post-editing, while the raw LLM output (zero-budget) remains usable. To ensure reproducibility to other languages, we made the code available at https://github.com/Tahaalshatiri/AutoArabic.
%U https://aclanthology.org/2025.arabicnlp-main.23/
%P 288-297
Markdown (Informal)
[AutoArabic: A Three-Stage Framework for Localizing Video-Text Retrieval Benchmarks](https://aclanthology.org/2025.arabicnlp-main.23/) (Eltahir et al., ArabicNLP 2025)
ACL