@inproceedings{salieb-etal-2026-nlpbeasts,
title = "{NLPB}easts at {A}ra{S}eg Shared Task 2026: A Character-Word {B}i{LSTM}-{CRF} for Paragraph-Aware {A}rabic Sentence Segmentation",
author = "Salieb, Michael Reda and
Aziz, Andrew and
Soliman, Kyrollos Zakaria and
Shaker, Mina Nady Serbana",
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: Shared Tasks",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.arabicnlp-sharedtasks.115/",
pages = "933--937",
ISBN = "979-8-89176-497-2",
abstract = "We describe NLPBeasts' submission to the Paragraph-Aware (PA) Closed track of the AraSeg 2026 shared task, which frames Arabic sentence segmentation as per-token binary boundary classification. We compare two closed-track-compliant systems trained with out any pretrained embeddings or language models: (1) a feature-based CRF using handcrafted punctuation, brackets, and paragraphs distance features, and (2) a character- and word-level BiLSTM with a linear-chain CRF decoder, both trained entirely from scratch on the shared task{'}s training and development splits. The feature-based CRF plateaus at a hard recall ceiling ({\ensuremath{\approx}}82{--}83{\%}) regardless of regularization strength, while the BiLSTM-CRF captures substantially more boundary cases through learned context. Our BiLSTM-CRF system, using only permitted training data with the development split reserved strictly for model selection, achieves P=86.37, R=90.51, and F1=87.60 on the blind test set. We additionally report ablations on chunk length and training hardware (CPU vs. GPU), neither of which meaningfully changes development F1, suggesting our system{'}s performance ceiling on this closed-track data size is data-limited rather than compute- or context-length-limited."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="salieb-etal-2026-nlpbeasts">
<titleInfo>
<title>NLPBeasts at AraSeg Shared Task 2026: A Character-Word BiLSTM-CRF for Paragraph-Aware Arabic Sentence Segmentation</title>
</titleInfo>
<name type="personal">
<namePart type="given">Michael</namePart>
<namePart type="given">Reda</namePart>
<namePart type="family">Salieb</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Andrew</namePart>
<namePart type="family">Aziz</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kyrollos</namePart>
<namePart type="given">Zakaria</namePart>
<namePart type="family">Soliman</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mina</namePart>
<namePart type="given">Nady</namePart>
<namePart type="given">Serbana</namePart>
<namePart type="family">Shaker</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: Shared Tasks</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">979-8-89176-497-2</identifier>
</relatedItem>
<abstract>We describe NLPBeasts’ submission to the Paragraph-Aware (PA) Closed track of the AraSeg 2026 shared task, which frames Arabic sentence segmentation as per-token binary boundary classification. We compare two closed-track-compliant systems trained with out any pretrained embeddings or language models: (1) a feature-based CRF using handcrafted punctuation, brackets, and paragraphs distance features, and (2) a character- and word-level BiLSTM with a linear-chain CRF decoder, both trained entirely from scratch on the shared task’s training and development splits. The feature-based CRF plateaus at a hard recall ceiling (\ensuremath\approx82–83%) regardless of regularization strength, while the BiLSTM-CRF captures substantially more boundary cases through learned context. Our BiLSTM-CRF system, using only permitted training data with the development split reserved strictly for model selection, achieves P=86.37, R=90.51, and F1=87.60 on the blind test set. We additionally report ablations on chunk length and training hardware (CPU vs. GPU), neither of which meaningfully changes development F1, suggesting our system’s performance ceiling on this closed-track data size is data-limited rather than compute- or context-length-limited.</abstract>
<identifier type="citekey">salieb-etal-2026-nlpbeasts</identifier>
<location>
<url>https://aclanthology.org/2026.arabicnlp-sharedtasks.115/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>933</start>
<end>937</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T NLPBeasts at AraSeg Shared Task 2026: A Character-Word BiLSTM-CRF for Paragraph-Aware Arabic Sentence Segmentation
%A Salieb, Michael Reda
%A Aziz, Andrew
%A Soliman, Kyrollos Zakaria
%A Shaker, Mina Nady Serbana
%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: Shared Tasks
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-497-2
%F salieb-etal-2026-nlpbeasts
%X We describe NLPBeasts’ submission to the Paragraph-Aware (PA) Closed track of the AraSeg 2026 shared task, which frames Arabic sentence segmentation as per-token binary boundary classification. We compare two closed-track-compliant systems trained with out any pretrained embeddings or language models: (1) a feature-based CRF using handcrafted punctuation, brackets, and paragraphs distance features, and (2) a character- and word-level BiLSTM with a linear-chain CRF decoder, both trained entirely from scratch on the shared task’s training and development splits. The feature-based CRF plateaus at a hard recall ceiling (\ensuremath\approx82–83%) regardless of regularization strength, while the BiLSTM-CRF captures substantially more boundary cases through learned context. Our BiLSTM-CRF system, using only permitted training data with the development split reserved strictly for model selection, achieves P=86.37, R=90.51, and F1=87.60 on the blind test set. We additionally report ablations on chunk length and training hardware (CPU vs. GPU), neither of which meaningfully changes development F1, suggesting our system’s performance ceiling on this closed-track data size is data-limited rather than compute- or context-length-limited.
%U https://aclanthology.org/2026.arabicnlp-sharedtasks.115/
%P 933-937
Markdown (Informal)
[NLPBeasts at AraSeg Shared Task 2026: A Character-Word BiLSTM-CRF for Paragraph-Aware Arabic Sentence Segmentation](https://aclanthology.org/2026.arabicnlp-sharedtasks.115/) (Salieb et al., ArabicNLP 2026)
ACL