@inproceedings{rossini-van-der-plas-2026-124,
title = "From 124 Million Tokens to 1,021 Neologisms: A Large-Scale Pipeline for Automatic Neologism Detection",
author = "Rossini, Diego and
van der Plas, Lonneke",
editor = "Oleskeviciene, Giedre Valunaite and
Giouli, Voula and
Armaselu, Florentina and
Liebeskind, Chaya and
McGillivray, Barbara",
booktitle = "Proceedings of the Workshop Neology and Large Language Models",
month = may,
year = "2026",
address = "Palma, Mallorca (Spain)",
publisher = "ELRA Language Resources Association (ELRA)",
url = "https://aclanthology.org/2026.neollm-1.1/",
doi = "10.63317/4o6ks86o293r",
pages = "1--15",
abstract = "We present a scalable, modular pipeline for automatic neologism detection that combines rule-based filtering with LLM classification. The pipeline is grounded in two complementary word-formation frameworks, grammatical and extra-grammatical morphology, which jointly define the scope of what counts as a neologism and inform a four-class classification scheme (NEOLOGISM, ENTITY, FOREIGN, NONE). While designed to be modular and transferable at the architectural level, the pipeline is instantiated on 527 million English-language Reddit posts spanning 2005{--}2024. From this corpus, we extract 124.6 million unique tokens and reduce them by over 99.99{\%} to yield 1,021 neologism candidates, a set small enough for manual expert verification. Multiple LLMs independently classify each candidate via majority vote, with a final verification step, revealing substantial cross-model disagreement and highlighting the challenge of operationalizing neologism detection at scale. Manual annotation of all 1,021 candidates confirms that 599 (58.7{\%}) are genuine lexical innovations."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="rossini-van-der-plas-2026-124">
<titleInfo>
<title>From 124 Million Tokens to 1,021 Neologisms: A Large-Scale Pipeline for Automatic Neologism Detection</title>
</titleInfo>
<name type="personal">
<namePart type="given">Diego</namePart>
<namePart type="family">Rossini</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Lonneke</namePart>
<namePart type="family">van der Plas</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>Proceedings of the Workshop Neology and Large Language Models</title>
</titleInfo>
<name type="personal">
<namePart type="given">Giedre</namePart>
<namePart type="given">Valunaite</namePart>
<namePart type="family">Oleskeviciene</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Voula</namePart>
<namePart type="family">Giouli</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Florentina</namePart>
<namePart type="family">Armaselu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chaya</namePart>
<namePart type="family">Liebeskind</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Barbara</namePart>
<namePart type="family">McGillivray</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>ELRA Language Resources Association (ELRA)</publisher>
<place>
<placeTerm type="text">Palma, Mallorca (Spain)</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>We present a scalable, modular pipeline for automatic neologism detection that combines rule-based filtering with LLM classification. The pipeline is grounded in two complementary word-formation frameworks, grammatical and extra-grammatical morphology, which jointly define the scope of what counts as a neologism and inform a four-class classification scheme (NEOLOGISM, ENTITY, FOREIGN, NONE). While designed to be modular and transferable at the architectural level, the pipeline is instantiated on 527 million English-language Reddit posts spanning 2005–2024. From this corpus, we extract 124.6 million unique tokens and reduce them by over 99.99% to yield 1,021 neologism candidates, a set small enough for manual expert verification. Multiple LLMs independently classify each candidate via majority vote, with a final verification step, revealing substantial cross-model disagreement and highlighting the challenge of operationalizing neologism detection at scale. Manual annotation of all 1,021 candidates confirms that 599 (58.7%) are genuine lexical innovations.</abstract>
<identifier type="citekey">rossini-van-der-plas-2026-124</identifier>
<identifier type="doi">10.63317/4o6ks86o293r</identifier>
<location>
<url>https://aclanthology.org/2026.neollm-1.1/</url>
</location>
<part>
<date>2026-05</date>
<extent unit="page">
<start>1</start>
<end>15</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T From 124 Million Tokens to 1,021 Neologisms: A Large-Scale Pipeline for Automatic Neologism Detection
%A Rossini, Diego
%A van der Plas, Lonneke
%Y Oleskeviciene, Giedre Valunaite
%Y Giouli, Voula
%Y Armaselu, Florentina
%Y Liebeskind, Chaya
%Y McGillivray, Barbara
%S Proceedings of the Workshop Neology and Large Language Models
%D 2026
%8 May
%I ELRA Language Resources Association (ELRA)
%C Palma, Mallorca (Spain)
%F rossini-van-der-plas-2026-124
%X We present a scalable, modular pipeline for automatic neologism detection that combines rule-based filtering with LLM classification. The pipeline is grounded in two complementary word-formation frameworks, grammatical and extra-grammatical morphology, which jointly define the scope of what counts as a neologism and inform a four-class classification scheme (NEOLOGISM, ENTITY, FOREIGN, NONE). While designed to be modular and transferable at the architectural level, the pipeline is instantiated on 527 million English-language Reddit posts spanning 2005–2024. From this corpus, we extract 124.6 million unique tokens and reduce them by over 99.99% to yield 1,021 neologism candidates, a set small enough for manual expert verification. Multiple LLMs independently classify each candidate via majority vote, with a final verification step, revealing substantial cross-model disagreement and highlighting the challenge of operationalizing neologism detection at scale. Manual annotation of all 1,021 candidates confirms that 599 (58.7%) are genuine lexical innovations.
%R 10.63317/4o6ks86o293r
%U https://aclanthology.org/2026.neollm-1.1/
%U https://doi.org/10.63317/4o6ks86o293r
%P 1-15
Markdown (Informal)
[From 124 Million Tokens to 1,021 Neologisms: A Large-Scale Pipeline for Automatic Neologism Detection](https://aclanthology.org/2026.neollm-1.1/) (Rossini & van der Plas, NeoLLM 2026)
ACL