@inproceedings{ahmadi-sennrich-2026-language,
title = "Language Identification for Non-Standard Varieties: Language Varieties as Geolects on a Geodesic Grid",
author = "Ahmadi, Sina and
Sennrich, Rico",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.emnlp-main.2175/",
pages = "49459--49481",
ISBN = "979-8-89176-459-0",
abstract = "Language identification is a critical preprocessing step for curating multilingual corpora. While current approaches treat language varieties as clearly bounded categories, such discrete labels obscure the continuous nature of linguistic variation in practice. We propose reframing non-standard variety identification as a geographically-grounded task, representing varieties as geolects, that is, linguistic data anchored to geographic coordinates. Rather than classifying texts into predefined labels, we place them on a geodesic grid, where predictions are points on the earth{'}s surface and variety distinctions emerge from spatial structure. We introduce GeoVar, a fine-grained dataset of 517,985 samples spanning 929 geolects across 20 languages and seven dialect continua in Europe and the Middle East, and develop a grid-based framework with three geolect-to-cell mapping strategies of increasing spatial smoothness: centroid, uniform, and Gaussian. We compare various geolocation methods, with the best model reaching 0.97 accuracy on compact continua such as Basque and Caspian. Accuracy decreases with the geographic dispersion of varieties, with North Germanic being the hardest case. Our model handles non-standard variety text more reliably than current LID tools, while additionally operating at sub-language granularity."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="ahmadi-sennrich-2026-language">
<titleInfo>
<title>Language Identification for Non-Standard Varieties: Language Varieties as Geolects on a Geodesic Grid</title>
</titleInfo>
<name type="personal">
<namePart type="given">Sina</namePart>
<namePart type="family">Ahmadi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Rico</namePart>
<namePart type="family">Sennrich</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 2026 Conference on Empirical Methods in Natural Language Processing</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hung-Yi</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sunipa</namePart>
<namePart type="family">Dev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jessy</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Desmond</namePart>
<namePart type="family">Elliott</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">979-8-89176-459-0</identifier>
</relatedItem>
<abstract>Language identification is a critical preprocessing step for curating multilingual corpora. While current approaches treat language varieties as clearly bounded categories, such discrete labels obscure the continuous nature of linguistic variation in practice. We propose reframing non-standard variety identification as a geographically-grounded task, representing varieties as geolects, that is, linguistic data anchored to geographic coordinates. Rather than classifying texts into predefined labels, we place them on a geodesic grid, where predictions are points on the earth’s surface and variety distinctions emerge from spatial structure. We introduce GeoVar, a fine-grained dataset of 517,985 samples spanning 929 geolects across 20 languages and seven dialect continua in Europe and the Middle East, and develop a grid-based framework with three geolect-to-cell mapping strategies of increasing spatial smoothness: centroid, uniform, and Gaussian. We compare various geolocation methods, with the best model reaching 0.97 accuracy on compact continua such as Basque and Caspian. Accuracy decreases with the geographic dispersion of varieties, with North Germanic being the hardest case. Our model handles non-standard variety text more reliably than current LID tools, while additionally operating at sub-language granularity.</abstract>
<identifier type="citekey">ahmadi-sennrich-2026-language</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-main.2175/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>49459</start>
<end>49481</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Language Identification for Non-Standard Varieties: Language Varieties as Geolects on a Geodesic Grid
%A Ahmadi, Sina
%A Sennrich, Rico
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-459-0
%F ahmadi-sennrich-2026-language
%X Language identification is a critical preprocessing step for curating multilingual corpora. While current approaches treat language varieties as clearly bounded categories, such discrete labels obscure the continuous nature of linguistic variation in practice. We propose reframing non-standard variety identification as a geographically-grounded task, representing varieties as geolects, that is, linguistic data anchored to geographic coordinates. Rather than classifying texts into predefined labels, we place them on a geodesic grid, where predictions are points on the earth’s surface and variety distinctions emerge from spatial structure. We introduce GeoVar, a fine-grained dataset of 517,985 samples spanning 929 geolects across 20 languages and seven dialect continua in Europe and the Middle East, and develop a grid-based framework with three geolect-to-cell mapping strategies of increasing spatial smoothness: centroid, uniform, and Gaussian. We compare various geolocation methods, with the best model reaching 0.97 accuracy on compact continua such as Basque and Caspian. Accuracy decreases with the geographic dispersion of varieties, with North Germanic being the hardest case. Our model handles non-standard variety text more reliably than current LID tools, while additionally operating at sub-language granularity.
%U https://aclanthology.org/2026.emnlp-main.2175/
%P 49459-49481
Markdown (Informal)
[Language Identification for Non-Standard Varieties: Language Varieties as Geolects on a Geodesic Grid](https://aclanthology.org/2026.emnlp-main.2175/) (Ahmadi & Sennrich, EMNLP 2026)
ACL