@inproceedings{fetouh-etal-2026-reglat,
title = "{REGLAT} at {A}bjad{M}ed: Handling Imbalanced {A}rabic Medical Text Classification via Hierarchical {KNN}-{MLP} Architecture",
author = "Fetouh, Ahmed Megahed and
Rahmath, Mohammed and
Dawood, Omer and
Labib, Mariam and
Ashraf, Nsrin and
Nayel, Hamada",
booktitle = "Proceedings of the 2nd Workshop on {NLP} for Languages Using {A}rabic Script",
month = mar,
year = "2026",
address = "Rabat, Morocco",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.abjadnlp-1.46/",
pages = "394--397",
abstract = "In this paper, we demonstrate the system submitted to the shared task of medical text classification in Arabic. We proposed a single-model approach based on fine-tuned LLM-based embedding combined with hierarchical classical classifiers, achieving a competitive macro F1-score of 0.46 on the blind test set. We explored various modeling strategies, including tree-based ensembles, LLM, and hierarchical correction for rare classes, highlighting the effectiveness of domain-specific fine-tuning in low-resource settings. The results demonstrate that a single fine-tuned Arabic BERT variant can serve as a strong baseline in extreme imbalance scenarios, outperforming more complex ensembles in simplicity and reproducibility."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="fetouh-etal-2026-reglat">
<titleInfo>
<title>REGLAT at AbjadMed: Handling Imbalanced Arabic Medical Text Classification via Hierarchical KNN-MLP Architecture</title>
</titleInfo>
<name type="personal">
<namePart type="given">Ahmed</namePart>
<namePart type="given">Megahed</namePart>
<namePart type="family">Fetouh</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mohammed</namePart>
<namePart type="family">Rahmath</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Omer</namePart>
<namePart type="family">Dawood</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mariam</namePart>
<namePart type="family">Labib</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Nsrin</namePart>
<namePart type="family">Ashraf</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hamada</namePart>
<namePart type="family">Nayel</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-03</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 2nd Workshop on NLP for Languages Using Arabic Script</title>
</titleInfo>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Rabat, Morocco</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>In this paper, we demonstrate the system submitted to the shared task of medical text classification in Arabic. We proposed a single-model approach based on fine-tuned LLM-based embedding combined with hierarchical classical classifiers, achieving a competitive macro F1-score of 0.46 on the blind test set. We explored various modeling strategies, including tree-based ensembles, LLM, and hierarchical correction for rare classes, highlighting the effectiveness of domain-specific fine-tuning in low-resource settings. The results demonstrate that a single fine-tuned Arabic BERT variant can serve as a strong baseline in extreme imbalance scenarios, outperforming more complex ensembles in simplicity and reproducibility.</abstract>
<identifier type="citekey">fetouh-etal-2026-reglat</identifier>
<location>
<url>https://aclanthology.org/2026.abjadnlp-1.46/</url>
</location>
<part>
<date>2026-03</date>
<extent unit="page">
<start>394</start>
<end>397</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T REGLAT at AbjadMed: Handling Imbalanced Arabic Medical Text Classification via Hierarchical KNN-MLP Architecture
%A Fetouh, Ahmed Megahed
%A Rahmath, Mohammed
%A Dawood, Omer
%A Labib, Mariam
%A Ashraf, Nsrin
%A Nayel, Hamada
%S Proceedings of the 2nd Workshop on NLP for Languages Using Arabic Script
%D 2026
%8 March
%I Association for Computational Linguistics
%C Rabat, Morocco
%F fetouh-etal-2026-reglat
%X In this paper, we demonstrate the system submitted to the shared task of medical text classification in Arabic. We proposed a single-model approach based on fine-tuned LLM-based embedding combined with hierarchical classical classifiers, achieving a competitive macro F1-score of 0.46 on the blind test set. We explored various modeling strategies, including tree-based ensembles, LLM, and hierarchical correction for rare classes, highlighting the effectiveness of domain-specific fine-tuning in low-resource settings. The results demonstrate that a single fine-tuned Arabic BERT variant can serve as a strong baseline in extreme imbalance scenarios, outperforming more complex ensembles in simplicity and reproducibility.
%U https://aclanthology.org/2026.abjadnlp-1.46/
%P 394-397
Markdown (Informal)
[REGLAT at AbjadMed: Handling Imbalanced Arabic Medical Text Classification via Hierarchical KNN-MLP Architecture](https://aclanthology.org/2026.abjadnlp-1.46/) (Fetouh et al., AbjadNLP 2026)
ACL