@inproceedings{xie-etal-2026-agentictagger,
title = "{A}gentic{T}agger: Generating Structured Item Representation for Recommendation with {LLM} Agents",
author = "Xie, Zhouhang and
Peng, Bo and
He, Zhankui and
Chen, Ziqi and
Han, Yizhao and
Ye, Isabella and
Coleman, Benjamin and
Sachdeva, Noveen and
Pereira, Fernando and
McAuley, Julian and
Kang, Wang-Cheng and
Cheng, Derek Zhiyuan and
Wang, Beidou and
Brown, Randolph G.",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {EMNLP} 2026",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-emnlp.2069/",
pages = "44031--44051",
ISBN = "979-8-89176-462-0",
abstract = "High-quality representations are a core requirement for effective recommendation. In this work, we study the problem of LLM-based descriptor generation, i.e., generating keyphrase-like natural language item representations with minimal constraints on downstream applications. However, open-ended generation provides little control over the generation space, leading to high cardinality, low-performance descriptors that render downstream modeling challenging. To this end, we propose AgenticTagger, a framework that queries LLMs to represent items with text descriptors drawn from an automatically mined vocabulary.\par AgenticTagger features two core stages: (1) a vocabulary-building stage in which a set of hierarchical, low-cardinality, and high-quality descriptors is identified, and (2) a vocabulary-assignment stage in which LLMs assign in-vocabulary descriptors to items. To ensure the vocabulary effectively describes the item corpus of interest, we design a multi-agent reflection mechanism in which an architect LLM iteratively refines the vocabulary guided by parallelized feedback from annotator LLMs that validate the vocabulary against item data. Experiments on public and private datasets show AgenticTagger brings consistent improvements across diverse recommendation scenarios, including generative retrieval, ranking, and critique-based recommendation."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="xie-etal-2026-agentictagger">
<titleInfo>
<title>AgenticTagger: Generating Structured Item Representation for Recommendation with LLM Agents</title>
</titleInfo>
<name type="personal">
<namePart type="given">Zhouhang</namePart>
<namePart type="family">Xie</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Bo</namePart>
<namePart type="family">Peng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhankui</namePart>
<namePart type="family">He</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ziqi</namePart>
<namePart type="family">Chen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yizhao</namePart>
<namePart type="family">Han</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Isabella</namePart>
<namePart type="family">Ye</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Benjamin</namePart>
<namePart type="family">Coleman</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Noveen</namePart>
<namePart type="family">Sachdeva</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Fernando</namePart>
<namePart type="family">Pereira</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Julian</namePart>
<namePart type="family">McAuley</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Wang-Cheng</namePart>
<namePart type="family">Kang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Derek</namePart>
<namePart type="given">Zhiyuan</namePart>
<namePart type="family">Cheng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Beidou</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Randolph</namePart>
<namePart type="given">G</namePart>
<namePart type="family">Brown</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>Findings of the Association for Computational Linguistics: EMNLP 2026</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">979-8-89176-462-0</identifier>
</relatedItem>
<abstract>High-quality representations are a core requirement for effective recommendation. In this work, we study the problem of LLM-based descriptor generation, i.e., generating keyphrase-like natural language item representations with minimal constraints on downstream applications. However, open-ended generation provides little control over the generation space, leading to high cardinality, low-performance descriptors that render downstream modeling challenging. To this end, we propose AgenticTagger, a framework that queries LLMs to represent items with text descriptors drawn from an automatically mined vocabulary.\par AgenticTagger features two core stages: (1) a vocabulary-building stage in which a set of hierarchical, low-cardinality, and high-quality descriptors is identified, and (2) a vocabulary-assignment stage in which LLMs assign in-vocabulary descriptors to items. To ensure the vocabulary effectively describes the item corpus of interest, we design a multi-agent reflection mechanism in which an architect LLM iteratively refines the vocabulary guided by parallelized feedback from annotator LLMs that validate the vocabulary against item data. Experiments on public and private datasets show AgenticTagger brings consistent improvements across diverse recommendation scenarios, including generative retrieval, ranking, and critique-based recommendation.</abstract>
<identifier type="citekey">xie-etal-2026-agentictagger</identifier>
<location>
<url>https://aclanthology.org/2026.findings-emnlp.2069/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>44031</start>
<end>44051</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T AgenticTagger: Generating Structured Item Representation for Recommendation with LLM Agents
%A Xie, Zhouhang
%A Peng, Bo
%A He, Zhankui
%A Chen, Ziqi
%A Han, Yizhao
%A Ye, Isabella
%A Coleman, Benjamin
%A Sachdeva, Noveen
%A Pereira, Fernando
%A McAuley, Julian
%A Kang, Wang-Cheng
%A Cheng, Derek Zhiyuan
%A Wang, Beidou
%A Brown, Randolph G.
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Findings of the Association for Computational Linguistics: EMNLP 2026
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-462-0
%F xie-etal-2026-agentictagger
%X High-quality representations are a core requirement for effective recommendation. In this work, we study the problem of LLM-based descriptor generation, i.e., generating keyphrase-like natural language item representations with minimal constraints on downstream applications. However, open-ended generation provides little control over the generation space, leading to high cardinality, low-performance descriptors that render downstream modeling challenging. To this end, we propose AgenticTagger, a framework that queries LLMs to represent items with text descriptors drawn from an automatically mined vocabulary.\par AgenticTagger features two core stages: (1) a vocabulary-building stage in which a set of hierarchical, low-cardinality, and high-quality descriptors is identified, and (2) a vocabulary-assignment stage in which LLMs assign in-vocabulary descriptors to items. To ensure the vocabulary effectively describes the item corpus of interest, we design a multi-agent reflection mechanism in which an architect LLM iteratively refines the vocabulary guided by parallelized feedback from annotator LLMs that validate the vocabulary against item data. Experiments on public and private datasets show AgenticTagger brings consistent improvements across diverse recommendation scenarios, including generative retrieval, ranking, and critique-based recommendation.
%U https://aclanthology.org/2026.findings-emnlp.2069/
%P 44031-44051
Markdown (Informal)
[AgenticTagger: Generating Structured Item Representation for Recommendation with LLM Agents](https://aclanthology.org/2026.findings-emnlp.2069/) (Xie et al., Findings 2026)
ACL
- Zhouhang Xie, Bo Peng, Zhankui He, Ziqi Chen, Yizhao Han, Isabella Ye, Benjamin Coleman, Noveen Sachdeva, Fernando Pereira, Julian McAuley, Wang-Cheng Kang, Derek Zhiyuan Cheng, Beidou Wang, and Randolph G. Brown. 2026. AgenticTagger: Generating Structured Item Representation for Recommendation with LLM Agents. In Findings of the Association for Computational Linguistics: EMNLP 2026, pages 44031–44051, Budapest, Hungary. Association for Computational Linguistics.