@inproceedings{horych-etal-2025-promises,
title = "The Promises and Pitfalls of {LLM} Annotations in Dataset Labeling: a Case Study on Media Bias Detection",
author = "Horych, Tom{\'a}{\v{s}} and
Mandl, Christoph and
Ruas, Terry and
Greiner-Petter, Andre and
Gipp, Bela and
Aizawa, Akiko and
Spinde, Timo",
editor = "Chiruzzo, Luis and
Ritter, Alan and
Wang, Lu",
booktitle = "Findings of the Association for Computational Linguistics: NAACL 2025",
month = apr,
year = "2025",
address = "Albuquerque, New Mexico",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.findings-naacl.75/",
doi = "10.18653/v1/2025.findings-naacl.75",
pages = "1370--1386",
ISBN = "979-8-89176-195-7",
abstract = "High annotation costs from hiring or crowdsourcing complicate the creation of large, high-quality datasets needed for training reliable text classifiers. Recent research suggests using Large Language Models (LLMs) to automate the annotation process, reducing these costs while maintaining data quality. LLMs have shown promising results in annotating downstream tasks like hate speech detection and political framing. Building on the success in these areas, this study investigates whether LLMs are viable for annotating a complex task of media bias detection and whether a downstream media bias classifier can be trained on such data. We create Annolexical, the first large-scale dataset for media bias classification with over 48k synthetically annotated examples. Our classifier fine-tuned on it surpasses all of the annotator LLMs by 5-9{\%} in Mathew{'}s Correlation Coefficient (MCC) and performs close to or outperforms the model trained on human-labeled data when evaluated on two media bias benchmark datasets (BABE and BASIL). This study demonstrates how our approach significantly reduces the cost of dataset creation in the media bias domain and, by extension - the development of the classifiers, while our subsequent behavioral stress-testing reveals some of its current limitations and trade-offs."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="horych-etal-2025-promises">
<titleInfo>
<title>The Promises and Pitfalls of LLM Annotations in Dataset Labeling: a Case Study on Media Bias Detection</title>
</titleInfo>
<name type="personal">
<namePart type="given">Tomáš</namePart>
<namePart type="family">Horych</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Christoph</namePart>
<namePart type="family">Mandl</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Terry</namePart>
<namePart type="family">Ruas</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Andre</namePart>
<namePart type="family">Greiner-Petter</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Bela</namePart>
<namePart type="family">Gipp</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Akiko</namePart>
<namePart type="family">Aizawa</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Timo</namePart>
<namePart type="family">Spinde</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-04</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: NAACL 2025</title>
</titleInfo>
<name type="personal">
<namePart type="given">Luis</namePart>
<namePart type="family">Chiruzzo</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Alan</namePart>
<namePart type="family">Ritter</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Lu</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Albuquerque, New Mexico</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-195-7</identifier>
</relatedItem>
<abstract>High annotation costs from hiring or crowdsourcing complicate the creation of large, high-quality datasets needed for training reliable text classifiers. Recent research suggests using Large Language Models (LLMs) to automate the annotation process, reducing these costs while maintaining data quality. LLMs have shown promising results in annotating downstream tasks like hate speech detection and political framing. Building on the success in these areas, this study investigates whether LLMs are viable for annotating a complex task of media bias detection and whether a downstream media bias classifier can be trained on such data. We create Annolexical, the first large-scale dataset for media bias classification with over 48k synthetically annotated examples. Our classifier fine-tuned on it surpasses all of the annotator LLMs by 5-9% in Mathew’s Correlation Coefficient (MCC) and performs close to or outperforms the model trained on human-labeled data when evaluated on two media bias benchmark datasets (BABE and BASIL). This study demonstrates how our approach significantly reduces the cost of dataset creation in the media bias domain and, by extension - the development of the classifiers, while our subsequent behavioral stress-testing reveals some of its current limitations and trade-offs.</abstract>
<identifier type="citekey">horych-etal-2025-promises</identifier>
<identifier type="doi">10.18653/v1/2025.findings-naacl.75</identifier>
<location>
<url>https://aclanthology.org/2025.findings-naacl.75/</url>
</location>
<part>
<date>2025-04</date>
<extent unit="page">
<start>1370</start>
<end>1386</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T The Promises and Pitfalls of LLM Annotations in Dataset Labeling: a Case Study on Media Bias Detection
%A Horych, Tomáš
%A Mandl, Christoph
%A Ruas, Terry
%A Greiner-Petter, Andre
%A Gipp, Bela
%A Aizawa, Akiko
%A Spinde, Timo
%Y Chiruzzo, Luis
%Y Ritter, Alan
%Y Wang, Lu
%S Findings of the Association for Computational Linguistics: NAACL 2025
%D 2025
%8 April
%I Association for Computational Linguistics
%C Albuquerque, New Mexico
%@ 979-8-89176-195-7
%F horych-etal-2025-promises
%X High annotation costs from hiring or crowdsourcing complicate the creation of large, high-quality datasets needed for training reliable text classifiers. Recent research suggests using Large Language Models (LLMs) to automate the annotation process, reducing these costs while maintaining data quality. LLMs have shown promising results in annotating downstream tasks like hate speech detection and political framing. Building on the success in these areas, this study investigates whether LLMs are viable for annotating a complex task of media bias detection and whether a downstream media bias classifier can be trained on such data. We create Annolexical, the first large-scale dataset for media bias classification with over 48k synthetically annotated examples. Our classifier fine-tuned on it surpasses all of the annotator LLMs by 5-9% in Mathew’s Correlation Coefficient (MCC) and performs close to or outperforms the model trained on human-labeled data when evaluated on two media bias benchmark datasets (BABE and BASIL). This study demonstrates how our approach significantly reduces the cost of dataset creation in the media bias domain and, by extension - the development of the classifiers, while our subsequent behavioral stress-testing reveals some of its current limitations and trade-offs.
%R 10.18653/v1/2025.findings-naacl.75
%U https://aclanthology.org/2025.findings-naacl.75/
%U https://doi.org/10.18653/v1/2025.findings-naacl.75
%P 1370-1386
Markdown (Informal)
[The Promises and Pitfalls of LLM Annotations in Dataset Labeling: a Case Study on Media Bias Detection](https://aclanthology.org/2025.findings-naacl.75/) (Horych et al., Findings 2025)
ACL