@inproceedings{zhang-etal-2025-leveraging-unpaired,
title = "Leveraging Unpaired Feedback for Long-Term {LLM}-based Recommendation Tuning",
author = "Zhang, Jizhi and
Gao, Chongming and
Shi, Wentao and
Chen, Xin and
Wang, Jingang and
Cai, Xunliang and
Feng, Fuli",
editor = "Christodoulopoulos, Christos and
Chakraborty, Tanmoy and
Rose, Carolyn and
Peng, Violet",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {EMNLP} 2025",
month = nov,
year = "2025",
address = "Suzhou, China",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.findings-emnlp.1332/",
doi = "10.18653/v1/2025.findings-emnlp.1332",
pages = "24507--24521",
ISBN = "979-8-89176-335-7",
abstract = "Most recommender systems focus on short-term objectives such as click-through rate, often at the expense of long-term user satisfaction. This can lead to echo chambers, where users are repeatedly exposed to redundant content. While recent efforts integrate Large Language Models (LLMs) into recommendation, they typically inherit this short-sighted focus. In this work, we highlight unpaired feedback{---}implicit signals such as continued engagement (positive) or silent disengagement (negative) that lack explicit contrastive labels{---}as a key challenge for long-term recommendation. Effectively learning from such feedback is crucial for improving LLM-based recommenders in dynamic user environments. To this end, we propose ULRec (Unpaired Feedback for Long-Term LLM-based Recommendation Tuning), a simple framework that fine-tunes LLMs using both positive and negative unpaired feedback. ULRec leverages the KTO algorithm to incorporate these signals without requiring paired supervision. Despite its simplicity, ULRec consistently improves long-term recommendation performance, demonstrating the value of modeling unpaired user feedback."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="zhang-etal-2025-leveraging-unpaired">
<titleInfo>
<title>Leveraging Unpaired Feedback for Long-Term LLM-based Recommendation Tuning</title>
</titleInfo>
<name type="personal">
<namePart type="given">Jizhi</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chongming</namePart>
<namePart type="family">Gao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Wentao</namePart>
<namePart type="family">Shi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xin</namePart>
<namePart type="family">Chen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jingang</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xunliang</namePart>
<namePart type="family">Cai</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Fuli</namePart>
<namePart type="family">Feng</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>Most recommender systems focus on short-term objectives such as click-through rate, often at the expense of long-term user satisfaction. This can lead to echo chambers, where users are repeatedly exposed to redundant content. While recent efforts integrate Large Language Models (LLMs) into recommendation, they typically inherit this short-sighted focus. In this work, we highlight unpaired feedback—implicit signals such as continued engagement (positive) or silent disengagement (negative) that lack explicit contrastive labels—as a key challenge for long-term recommendation. Effectively learning from such feedback is crucial for improving LLM-based recommenders in dynamic user environments. To this end, we propose ULRec (Unpaired Feedback for Long-Term LLM-based Recommendation Tuning), a simple framework that fine-tunes LLMs using both positive and negative unpaired feedback. ULRec leverages the KTO algorithm to incorporate these signals without requiring paired supervision. Despite its simplicity, ULRec consistently improves long-term recommendation performance, demonstrating the value of modeling unpaired user feedback.</abstract>
<identifier type="citekey">zhang-etal-2025-leveraging-unpaired</identifier>
<identifier type="doi">10.18653/v1/2025.findings-emnlp.1332</identifier>
<location>
<url>https://aclanthology.org/2025.findings-emnlp.1332/</url>
</location>
<part>
<date>2025-11</date>
<extent unit="page">
<start>24507</start>
<end>24521</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Leveraging Unpaired Feedback for Long-Term LLM-based Recommendation Tuning
%A Zhang, Jizhi
%A Gao, Chongming
%A Shi, Wentao
%A Chen, Xin
%A Wang, Jingang
%A Cai, Xunliang
%A Feng, Fuli
%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 zhang-etal-2025-leveraging-unpaired
%X Most recommender systems focus on short-term objectives such as click-through rate, often at the expense of long-term user satisfaction. This can lead to echo chambers, where users are repeatedly exposed to redundant content. While recent efforts integrate Large Language Models (LLMs) into recommendation, they typically inherit this short-sighted focus. In this work, we highlight unpaired feedback—implicit signals such as continued engagement (positive) or silent disengagement (negative) that lack explicit contrastive labels—as a key challenge for long-term recommendation. Effectively learning from such feedback is crucial for improving LLM-based recommenders in dynamic user environments. To this end, we propose ULRec (Unpaired Feedback for Long-Term LLM-based Recommendation Tuning), a simple framework that fine-tunes LLMs using both positive and negative unpaired feedback. ULRec leverages the KTO algorithm to incorporate these signals without requiring paired supervision. Despite its simplicity, ULRec consistently improves long-term recommendation performance, demonstrating the value of modeling unpaired user feedback.
%R 10.18653/v1/2025.findings-emnlp.1332
%U https://aclanthology.org/2025.findings-emnlp.1332/
%U https://doi.org/10.18653/v1/2025.findings-emnlp.1332
%P 24507-24521
Markdown (Informal)
[Leveraging Unpaired Feedback for Long-Term LLM-based Recommendation Tuning](https://aclanthology.org/2025.findings-emnlp.1332/) (Zhang et al., Findings 2025)
ACL