@inproceedings{pal-etal-2026-disentangling,
title = "Disentangling Approaches to Conversation Disentanglement: Fine-Tune or Learn from Scratch?",
author = "Pal, Debaditya and
Leuski, Anton and
Artstein, Ron and
Traum, David and
Georgila, Kallirroi",
editor = "Piperidis, Stelios and
Bel, N{\'u}ria and
van den Heuvel, Henk and
Ide, Nancy and
Krek, Simon and
Toral, Antonio",
booktitle = "Proceedings of the Fifteenth Language Resources and Evaluation Conference",
month = may,
year = "2026",
address = "Palma de Mallorca, Spain",
publisher = "ELRA Language Resource Association",
url = "https://aclanthology.org/2026.lrec-1.229/",
doi = "10.63317/3frwendckp7g",
pages = "2927--2941",
abstract = "Conversation disentanglement is the process of segmenting a stream of messages or utterances into separate conversations or ``threads'' that can be more easily understood and processed. We compare the performance of GPT-4o and GPT-4o Mini with deep learning models built from scratch for this task. We show that, using the same amount of training data, out-of-the-box GPT-4o performs poorly, and fine-tuning GPT-4o Mini results in performance comparable to learning small-size models from scratch (based on standard hand-crafted features for this task), with performance reaching 74.4{\%} F1-score for prediction of links between messages and 45.3{\%} F1-score for prediction of perfectly matching conversations. However, the fine-tuned GPT-4o Mini model underperforms when compared to models that utilize complex structural information. We also provide a new method for detailed analysis of the successes and failures of our models, and a new visualization method."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="pal-etal-2026-disentangling">
<titleInfo>
<title>Disentangling Approaches to Conversation Disentanglement: Fine-Tune or Learn from Scratch?</title>
</titleInfo>
<name type="personal">
<namePart type="given">Debaditya</namePart>
<namePart type="family">Pal</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Anton</namePart>
<namePart type="family">Leuski</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ron</namePart>
<namePart type="family">Artstein</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">David</namePart>
<namePart type="family">Traum</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kallirroi</namePart>
<namePart type="family">Georgila</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 Fifteenth Language Resources and Evaluation Conference</title>
</titleInfo>
<name type="personal">
<namePart type="given">Stelios</namePart>
<namePart type="family">Piperidis</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Núria</namePart>
<namePart type="family">Bel</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Henk</namePart>
<namePart type="family">van den Heuvel</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Nancy</namePart>
<namePart type="family">Ide</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Simon</namePart>
<namePart type="family">Krek</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Antonio</namePart>
<namePart type="family">Toral</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>ELRA Language Resource Association</publisher>
<place>
<placeTerm type="text">Palma de Mallorca, Spain</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Conversation disentanglement is the process of segmenting a stream of messages or utterances into separate conversations or “threads” that can be more easily understood and processed. We compare the performance of GPT-4o and GPT-4o Mini with deep learning models built from scratch for this task. We show that, using the same amount of training data, out-of-the-box GPT-4o performs poorly, and fine-tuning GPT-4o Mini results in performance comparable to learning small-size models from scratch (based on standard hand-crafted features for this task), with performance reaching 74.4% F1-score for prediction of links between messages and 45.3% F1-score for prediction of perfectly matching conversations. However, the fine-tuned GPT-4o Mini model underperforms when compared to models that utilize complex structural information. We also provide a new method for detailed analysis of the successes and failures of our models, and a new visualization method.</abstract>
<identifier type="citekey">pal-etal-2026-disentangling</identifier>
<identifier type="doi">10.63317/3frwendckp7g</identifier>
<location>
<url>https://aclanthology.org/2026.lrec-1.229/</url>
</location>
<part>
<date>2026-05</date>
<extent unit="page">
<start>2927</start>
<end>2941</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Disentangling Approaches to Conversation Disentanglement: Fine-Tune or Learn from Scratch?
%A Pal, Debaditya
%A Leuski, Anton
%A Artstein, Ron
%A Traum, David
%A Georgila, Kallirroi
%Y Piperidis, Stelios
%Y Bel, Núria
%Y van den Heuvel, Henk
%Y Ide, Nancy
%Y Krek, Simon
%Y Toral, Antonio
%S Proceedings of the Fifteenth Language Resources and Evaluation Conference
%D 2026
%8 May
%I ELRA Language Resource Association
%C Palma de Mallorca, Spain
%F pal-etal-2026-disentangling
%X Conversation disentanglement is the process of segmenting a stream of messages or utterances into separate conversations or “threads” that can be more easily understood and processed. We compare the performance of GPT-4o and GPT-4o Mini with deep learning models built from scratch for this task. We show that, using the same amount of training data, out-of-the-box GPT-4o performs poorly, and fine-tuning GPT-4o Mini results in performance comparable to learning small-size models from scratch (based on standard hand-crafted features for this task), with performance reaching 74.4% F1-score for prediction of links between messages and 45.3% F1-score for prediction of perfectly matching conversations. However, the fine-tuned GPT-4o Mini model underperforms when compared to models that utilize complex structural information. We also provide a new method for detailed analysis of the successes and failures of our models, and a new visualization method.
%R 10.63317/3frwendckp7g
%U https://aclanthology.org/2026.lrec-1.229/
%U https://doi.org/10.63317/3frwendckp7g
%P 2927-2941
Markdown (Informal)
[Disentangling Approaches to Conversation Disentanglement: Fine-Tune or Learn from Scratch?](https://aclanthology.org/2026.lrec-1.229/) (Pal et al., LREC 2026)
ACL