@inproceedings{waldendorf-etal-2025-multilingual,
title = "Multilingual Data Filtering using Synthetic Data from Large Language Models",
author = "Waldendorf, Jonas and
Haddow, Barry and
Birch, Alexandra and
Klimaszewski, Mateusz",
editor = "Christodoulopoulos, Christos and
Chakraborty, Tanmoy and
Rose, Carolyn and
Peng, Violet",
booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2025",
month = nov,
year = "2025",
address = "Suzhou, China",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.findings-emnlp.495/",
doi = "10.18653/v1/2025.findings-emnlp.495",
pages = "9317--9334",
ISBN = "979-8-89176-335-7",
abstract = "Filtering data, particularly data scraped from the internet, has long been recognised as a means to improve model performance. Recent studies have shown that effective filters can be created by utilising Large Language Models (LLMs) to synthetically label data, which is then used to train smaller neural models for filtering purposes. However, this approach has been tested mainly in English. Our paper extends this approach to languages beyond English, including languages not officially supported by the LLM. We validate our results on the downstream task of NMT and demonstrate that our approach is effective at both filtering parallel text for translation quality and filtering for domain specificity. For training the filtering model, we experiment with two different objectives for finetuning pre-trained transformers, as well as an efficient approach based on \textit{n}-gram language models."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="waldendorf-etal-2025-multilingual">
<titleInfo>
<title>Multilingual Data Filtering using Synthetic Data from Large Language Models</title>
</titleInfo>
<name type="personal">
<namePart type="given">Jonas</namePart>
<namePart type="family">Waldendorf</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Barry</namePart>
<namePart type="family">Haddow</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Alexandra</namePart>
<namePart type="family">Birch</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mateusz</namePart>
<namePart type="family">Klimaszewski</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-11</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: EMNLP 2025</title>
</titleInfo>
<name type="personal">
<namePart type="given">Christos</namePart>
<namePart type="family">Christodoulopoulos</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tanmoy</namePart>
<namePart type="family">Chakraborty</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Carolyn</namePart>
<namePart type="family">Rose</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Violet</namePart>
<namePart type="family">Peng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Suzhou, China</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-335-7</identifier>
</relatedItem>
<abstract>Filtering data, particularly data scraped from the internet, has long been recognised as a means to improve model performance. Recent studies have shown that effective filters can be created by utilising Large Language Models (LLMs) to synthetically label data, which is then used to train smaller neural models for filtering purposes. However, this approach has been tested mainly in English. Our paper extends this approach to languages beyond English, including languages not officially supported by the LLM. We validate our results on the downstream task of NMT and demonstrate that our approach is effective at both filtering parallel text for translation quality and filtering for domain specificity. For training the filtering model, we experiment with two different objectives for finetuning pre-trained transformers, as well as an efficient approach based on n-gram language models.</abstract>
<identifier type="citekey">waldendorf-etal-2025-multilingual</identifier>
<identifier type="doi">10.18653/v1/2025.findings-emnlp.495</identifier>
<location>
<url>https://aclanthology.org/2025.findings-emnlp.495/</url>
</location>
<part>
<date>2025-11</date>
<extent unit="page">
<start>9317</start>
<end>9334</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Multilingual Data Filtering using Synthetic Data from Large Language Models
%A Waldendorf, Jonas
%A Haddow, Barry
%A Birch, Alexandra
%A Klimaszewski, Mateusz
%Y Christodoulopoulos, Christos
%Y Chakraborty, Tanmoy
%Y Rose, Carolyn
%Y Peng, Violet
%S Findings of the Association for Computational Linguistics: EMNLP 2025
%D 2025
%8 November
%I Association for Computational Linguistics
%C Suzhou, China
%@ 979-8-89176-335-7
%F waldendorf-etal-2025-multilingual
%X Filtering data, particularly data scraped from the internet, has long been recognised as a means to improve model performance. Recent studies have shown that effective filters can be created by utilising Large Language Models (LLMs) to synthetically label data, which is then used to train smaller neural models for filtering purposes. However, this approach has been tested mainly in English. Our paper extends this approach to languages beyond English, including languages not officially supported by the LLM. We validate our results on the downstream task of NMT and demonstrate that our approach is effective at both filtering parallel text for translation quality and filtering for domain specificity. For training the filtering model, we experiment with two different objectives for finetuning pre-trained transformers, as well as an efficient approach based on n-gram language models.
%R 10.18653/v1/2025.findings-emnlp.495
%U https://aclanthology.org/2025.findings-emnlp.495/
%U https://doi.org/10.18653/v1/2025.findings-emnlp.495
%P 9317-9334
Markdown (Informal)
[Multilingual Data Filtering using Synthetic Data from Large Language Models](https://aclanthology.org/2025.findings-emnlp.495/) (Waldendorf et al., Findings 2025)
ACL