@inproceedings{istaiteh-etal-2026-wer,
title = "What {WER} Hides: A Closer Look at {A}lgerian Code-Switched {ASR}",
author = "Istaiteh, Othman and
Toughrai, Yassine and
Smaili, Kamel and
Mdhaffar, Salima and
Langlois, David and
Est{\`e}ve, Yannick",
editor = "Abdul-Mageed, Muhammad and
Elsayed, Tamer and
Bouamor, Houda and
El-Beltagy, Samhaa and
Al-Khalifa, Hend and
Abdelali, Ahmed and
Elmadany, AbdelRahim and
El Mekki, Abdellah and
Darwish, Kareem and
Alkhereyf, Sakhar and
Luqman, Hamzah and
Khalifa, Salam and
Magdy, Samar and
Alyafeai, Zaid and
Hamed, Injy and
Zaghouani, Wajdi and
Zitouni, Imed",
booktitle = "Proceedings of The Fourth {A}rabic Natural Language Processing Conference",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.arabicnlp-main.26/",
pages = "373--383",
ISBN = "979-8-89176-473-6",
abstract = "Automatic speech recognition for Algerian social-media speech is hard because of dialectal variation, Arabic-French code-switching, and non-standard orthography, yet these difficulties are usually summarised by a single word error rate that hides where recognition fails and whether gains reach information-bearing content. We evaluate seven systems on roughly 30 minutes of Algerian political social-media speech, annotated by language variety as Algerian dialect, Modern Standard Arabic, code-switched, or French: a Whisper-large-v3 baseline, five Whisper variants fine-tuned on different combinations of broadcast, dialectal, and code-switched speech, and the prompted Qwen3-Omni model. The best system reduces overall WER from 60.1{\%} to 31.3{\%}, but still reaches 44.0{\%} on code-switched speech, and named entities remain harder than ordinary tokens, with fewer than half preserved. These results show that ASR difficulty on Algerian social-media speech cannot be read from overall WER alone; it must be measured separately by language variety and by whether named entities survive."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="istaiteh-etal-2026-wer">
<titleInfo>
<title>What WER Hides: A Closer Look at Algerian Code-Switched ASR</title>
</titleInfo>
<name type="personal">
<namePart type="given">Othman</namePart>
<namePart type="family">Istaiteh</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yassine</namePart>
<namePart type="family">Toughrai</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kamel</namePart>
<namePart type="family">Smaili</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Salima</namePart>
<namePart type="family">Mdhaffar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">David</namePart>
<namePart type="family">Langlois</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yannick</namePart>
<namePart type="family">Estève</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 Fourth Arabic Natural Language Processing Conference</title>
</titleInfo>
<name type="personal">
<namePart type="given">Muhammad</namePart>
<namePart type="family">Abdul-Mageed</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tamer</namePart>
<namePart type="family">Elsayed</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Houda</namePart>
<namePart type="family">Bouamor</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Samhaa</namePart>
<namePart type="family">El-Beltagy</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hend</namePart>
<namePart type="family">Al-Khalifa</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">AbdelRahim</namePart>
<namePart type="family">Elmadany</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Abdellah</namePart>
<namePart type="family">El Mekki</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<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">Sakhar</namePart>
<namePart type="family">Alkhereyf</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hamzah</namePart>
<namePart type="family">Luqman</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">Samar</namePart>
<namePart type="family">Magdy</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">Injy</namePart>
<namePart type="family">Hamed</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">Imed</namePart>
<namePart type="family">Zitouni</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">979-8-89176-473-6</identifier>
</relatedItem>
<abstract>Automatic speech recognition for Algerian social-media speech is hard because of dialectal variation, Arabic-French code-switching, and non-standard orthography, yet these difficulties are usually summarised by a single word error rate that hides where recognition fails and whether gains reach information-bearing content. We evaluate seven systems on roughly 30 minutes of Algerian political social-media speech, annotated by language variety as Algerian dialect, Modern Standard Arabic, code-switched, or French: a Whisper-large-v3 baseline, five Whisper variants fine-tuned on different combinations of broadcast, dialectal, and code-switched speech, and the prompted Qwen3-Omni model. The best system reduces overall WER from 60.1% to 31.3%, but still reaches 44.0% on code-switched speech, and named entities remain harder than ordinary tokens, with fewer than half preserved. These results show that ASR difficulty on Algerian social-media speech cannot be read from overall WER alone; it must be measured separately by language variety and by whether named entities survive.</abstract>
<identifier type="citekey">istaiteh-etal-2026-wer</identifier>
<location>
<url>https://aclanthology.org/2026.arabicnlp-main.26/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>373</start>
<end>383</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T What WER Hides: A Closer Look at Algerian Code-Switched ASR
%A Istaiteh, Othman
%A Toughrai, Yassine
%A Smaili, Kamel
%A Mdhaffar, Salima
%A Langlois, David
%A Estève, Yannick
%Y Abdul-Mageed, Muhammad
%Y Elsayed, Tamer
%Y Bouamor, Houda
%Y El-Beltagy, Samhaa
%Y Al-Khalifa, Hend
%Y Abdelali, Ahmed
%Y Elmadany, AbdelRahim
%Y El Mekki, Abdellah
%Y Darwish, Kareem
%Y Alkhereyf, Sakhar
%Y Luqman, Hamzah
%Y Khalifa, Salam
%Y Magdy, Samar
%Y Alyafeai, Zaid
%Y Hamed, Injy
%Y Zaghouani, Wajdi
%Y Zitouni, Imed
%S Proceedings of The Fourth Arabic Natural Language Processing Conference
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-473-6
%F istaiteh-etal-2026-wer
%X Automatic speech recognition for Algerian social-media speech is hard because of dialectal variation, Arabic-French code-switching, and non-standard orthography, yet these difficulties are usually summarised by a single word error rate that hides where recognition fails and whether gains reach information-bearing content. We evaluate seven systems on roughly 30 minutes of Algerian political social-media speech, annotated by language variety as Algerian dialect, Modern Standard Arabic, code-switched, or French: a Whisper-large-v3 baseline, five Whisper variants fine-tuned on different combinations of broadcast, dialectal, and code-switched speech, and the prompted Qwen3-Omni model. The best system reduces overall WER from 60.1% to 31.3%, but still reaches 44.0% on code-switched speech, and named entities remain harder than ordinary tokens, with fewer than half preserved. These results show that ASR difficulty on Algerian social-media speech cannot be read from overall WER alone; it must be measured separately by language variety and by whether named entities survive.
%U https://aclanthology.org/2026.arabicnlp-main.26/
%P 373-383
Markdown (Informal)
[What WER Hides: A Closer Look at Algerian Code-Switched ASR](https://aclanthology.org/2026.arabicnlp-main.26/) (Istaiteh et al., ArabicNLP 2026)
ACL
- Othman Istaiteh, Yassine Toughrai, Kamel Smaili, Salima Mdhaffar, David Langlois, and Yannick Estève. 2026. What WER Hides: A Closer Look at Algerian Code-Switched ASR. In Proceedings of The Fourth Arabic Natural Language Processing Conference, pages 373–383, Budapest, Hungary. Association for Computational Linguistics.