When Bigger Isn’t Better: Evaluating LLMs for Arabic Sentiment Analysis

Mohamed Ibrahim; Abdullah Makki; Youssef Barakat; Nour Samy; Sarah Alhumoud

doi:10.63317/2kimof4u6y8x

When Bigger Isn’t Better: Evaluating LLMs for Arabic Sentiment Analysis

Mohamed Ibrahim, Abdullah Makki, Youssef Barakat, Nour Samy, Sarah AlHumoud

Correct Metadata for

Use this form to create a GitHub issue with structured data describing the correction. You will need a GitHub account. Once you create that issue, the correction will be reviewed by a staff member.

⚠️ Mobile Users: Submitting this form to create a new issue will only work with github.com, not the GitHub Mobile app.

Important: The Anthology treat PDFs as authoritative. Please use this form only to correct data that is out of line with the PDF. See our corrections guidelines if you need to change the PDF.

Title Adjust the title. Retain tags such as <fixed-case>.

Authors Adjust author names and order to match the PDF.

Abstract Correct abstract if needed. Retain XML formatting tags such as <tex-math>. You may use ... for bold, ... for italic, ... for underline, <sc>...</sc> for small-caps, <tt>...<tt> for typewriter text, <url>...</url> for URLs, <a href=...> for hyperlinks, and <par/> for paragraph breaks.

Verification against PDF Ensure that the new title/authors match the snapshot below. (If there is no snapshot or it is too small, consult the external publication .)

Authors concatenated from the text boxes above:

ALL author names match the snapshot above—including middle initials, hyphens, and accents.

Abstract

This study evaluates the performance of a fine-tuned Arabic sentiment transformer (CAMeL-MSA) against eight large language models (LLMs). Using zero-shot prompting across six Arabic sentiment datasets, we compare a specialized, task-specific approach against generalized model capabilities. Results show that the fine-tuned baseline substantially outperformed all LLMs on five of the six datasets in both accuracy and Macro F1-score. While LLMs offer versatility, this comparison highlights the continued practical superiority of task-specific fine-tuning over zero-shot prompting.

Anthology ID:: 2026.osact-1.4
Volume:: The 7th Workshop on Open-Source Arabic Corpora and Processing Tools (OSACT7) with 5 Shared Tasks
Month:: May
Year:: 2026
Address:: Palma, Mallorca (Spain)
Editors:: Hend Al-Khalifa, Mo El-Haj, Saad Ezzini
Venues:: OSACT | WS
SIG:
Publisher:: Association for Computational Linguistics
Note:
Pages:: 35–39
Language:
External URL:: https://lrec.elra.info/lrec2026-ws-osact-04
DOI:: 10.63317/2kimof4u6y8x
Bibkey:
Cite (ACL):: Mohamed Ibrahim, Abdullah Makki, Youssef Barakat, Nour Samy, and Sarah AlHumoud. 2026. When Bigger Isn’t Better: Evaluating LLMs for Arabic Sentiment Analysis. In The 7th Workshop on Open-Source Arabic Corpora and Processing Tools (OSACT7) with 5 Shared Tasks, pages 35–39, Palma, Mallorca (Spain). Association for Computational Linguistics.
Cite (Informal):: When Bigger Isn’t Better: Evaluating LLMs for Arabic Sentiment Analysis (Ibrahim et al., OSACT 2026)
Copy Citation:

External Cite Search Fix data

Export citation

BibTeX
MODS XML
Endnote
Preformatted

@inproceedings{ibrahim-etal-2026-bigger,
    title = "When Bigger Isn{'}t Better: Evaluating {LLM}s for {A}rabic Sentiment Analysis",
    author = "Ibrahim, Mohamed  and
      Makki, Abdullah  and
      Barakat, Youssef  and
      Samy, Nour  and
      AlHumoud, Sarah",
    editor = "Al-Khalifa, Hend  and
      El-Haj, Mo  and
      Ezzini, Saad",
    booktitle = "The 7th Workshop on Open-Source {A}rabic Corpora and Processing Tools ({OSACT}7) with 5 Shared Tasks",
    month = may,
    year = "2026",
    address = "Palma, Mallorca (Spain)",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2026.osact-1.4/",
    doi = "10.63317/2kimof4u6y8x",
    pages = "35--39",
    abstract = "This study evaluates the performance of a fine-tuned Arabic sentiment transformer (CAMeL-MSA) against eight large language models (LLMs). Using zero-shot prompting across six Arabic sentiment datasets, we compare a specialized, task-specific approach against generalized model capabilities. Results show that the fine-tuned baseline substantially outperformed all LLMs on five of the six datasets in both accuracy and Macro F1-score. While LLMs offer versatility, this comparison highlights the continued practical superiority of task-specific fine-tuning over zero-shot prompting."
}

Download as File

<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="ibrahim-etal-2026-bigger">
    <titleInfo>
        <title>When Bigger Isn’t Better: Evaluating LLMs for Arabic Sentiment Analysis</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Mohamed</namePart>
        <namePart type="family">Ibrahim</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Abdullah</namePart>
        <namePart type="family">Makki</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Youssef</namePart>
        <namePart type="family">Barakat</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Nour</namePart>
        <namePart type="family">Samy</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Sarah</namePart>
        <namePart type="family">AlHumoud</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2026-05</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>The 7th Workshop on Open-Source Arabic Corpora and Processing Tools (OSACT7) with 5 Shared Tasks</title>
        </titleInfo>
        <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">Mo</namePart>
            <namePart type="family">El-Haj</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Saad</namePart>
            <namePart type="family">Ezzini</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Palma, Mallorca (Spain)</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
    </relatedItem>
    <abstract>This study evaluates the performance of a fine-tuned Arabic sentiment transformer (CAMeL-MSA) against eight large language models (LLMs). Using zero-shot prompting across six Arabic sentiment datasets, we compare a specialized, task-specific approach against generalized model capabilities. Results show that the fine-tuned baseline substantially outperformed all LLMs on five of the six datasets in both accuracy and Macro F1-score. While LLMs offer versatility, this comparison highlights the continued practical superiority of task-specific fine-tuning over zero-shot prompting.</abstract>
    <identifier type="citekey">ibrahim-etal-2026-bigger</identifier>
    <identifier type="doi">10.63317/2kimof4u6y8x</identifier>
    <location>
        <url>https://aclanthology.org/2026.osact-1.4/</url>
    </location>
    <part>
        <date>2026-05</date>
        <extent unit="page">
            <start>35</start>
            <end>39</end>
        </extent>
    </part>
</mods>
</modsCollection>

Download as File

%0 Conference Proceedings
%T When Bigger Isn’t Better: Evaluating LLMs for Arabic Sentiment Analysis
%A Ibrahim, Mohamed
%A Makki, Abdullah
%A Barakat, Youssef
%A Samy, Nour
%A AlHumoud, Sarah
%Y Al-Khalifa, Hend
%Y El-Haj, Mo
%Y Ezzini, Saad
%S The 7th Workshop on Open-Source Arabic Corpora and Processing Tools (OSACT7) with 5 Shared Tasks
%D 2026
%8 May
%I Association for Computational Linguistics
%C Palma, Mallorca (Spain)
%F ibrahim-etal-2026-bigger
%X This study evaluates the performance of a fine-tuned Arabic sentiment transformer (CAMeL-MSA) against eight large language models (LLMs). Using zero-shot prompting across six Arabic sentiment datasets, we compare a specialized, task-specific approach against generalized model capabilities. Results show that the fine-tuned baseline substantially outperformed all LLMs on five of the six datasets in both accuracy and Macro F1-score. While LLMs offer versatility, this comparison highlights the continued practical superiority of task-specific fine-tuning over zero-shot prompting.
%R 10.63317/2kimof4u6y8x
%U https://aclanthology.org/2026.osact-1.4/
%U https://doi.org/10.63317/2kimof4u6y8x
%P 35-39

Download as File

Markdown (Informal)

[When Bigger Isn’t Better: Evaluating LLMs for Arabic Sentiment Analysis](https://aclanthology.org/2026.osact-1.4/) (Ibrahim et al., OSACT 2026)

When Bigger Isn’t Better: Evaluating LLMs for Arabic Sentiment Analysis (Ibrahim et al., OSACT 2026)

ACL

Mohamed Ibrahim, Abdullah Makki, Youssef Barakat, Nour Samy, and Sarah AlHumoud. 2026. When Bigger Isn’t Better: Evaluating LLMs for Arabic Sentiment Analysis. In The 7th Workshop on Open-Source Arabic Corpora and Processing Tools (OSACT7) with 5 Shared Tasks, pages 35–39, Palma, Mallorca (Spain). Association for Computational Linguistics.