@inproceedings{choi-etal-2026-pgmem,
title = "{PGM}em: Tightly Coupled Persona{--}Memory Graph for Lifelong Personalized Agents",
author = "Choi, Wonjun and
Kim, Yerim and
Lee, Yukyung and
Yoon, Susik",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.emnlp-main.1891/",
pages = "43087--43112",
ISBN = "979-8-89176-459-0",
abstract = "Long-term personalized dialogue agents must track user preferences as their personas evolve. Existing memory systems organize past events well, but store personas as flat profiles detached from the events that justify them. This loose coupling leads to the memory{--}persona validity gap and the persona-aware retrieval gap. We propose PGMem, a heterogeneous persona-memory graph that connects event and persona nodes through typed provenance and evidence edges, keeping each persona signal traceable to the events that support or revise it. At retrieval time, PGMem expands from query-relevant seeds and ranks signals by evidential validity. Across three benchmarks with small language model backbones, PGMem consistently outperforms summary-based, persona-aware, graph-structured, and agentic memory baselines, and improves performance as the context grows. The source codes of PGMem are available at https://github.com/wonjunchoi23/pgmem/"
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="choi-etal-2026-pgmem">
<titleInfo>
<title>PGMem: Tightly Coupled Persona–Memory Graph for Lifelong Personalized Agents</title>
</titleInfo>
<name type="personal">
<namePart type="given">Wonjun</namePart>
<namePart type="family">Choi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yerim</namePart>
<namePart type="family">Kim</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yukyung</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Susik</namePart>
<namePart type="family">Yoon</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>Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing</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">979-8-89176-459-0</identifier>
</relatedItem>
<abstract>Long-term personalized dialogue agents must track user preferences as their personas evolve. Existing memory systems organize past events well, but store personas as flat profiles detached from the events that justify them. This loose coupling leads to the memory–persona validity gap and the persona-aware retrieval gap. We propose PGMem, a heterogeneous persona-memory graph that connects event and persona nodes through typed provenance and evidence edges, keeping each persona signal traceable to the events that support or revise it. At retrieval time, PGMem expands from query-relevant seeds and ranks signals by evidential validity. Across three benchmarks with small language model backbones, PGMem consistently outperforms summary-based, persona-aware, graph-structured, and agentic memory baselines, and improves performance as the context grows. The source codes of PGMem are available at https://github.com/wonjunchoi23/pgmem/</abstract>
<identifier type="citekey">choi-etal-2026-pgmem</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-main.1891/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>43087</start>
<end>43112</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T PGMem: Tightly Coupled Persona–Memory Graph for Lifelong Personalized Agents
%A Choi, Wonjun
%A Kim, Yerim
%A Lee, Yukyung
%A Yoon, Susik
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-459-0
%F choi-etal-2026-pgmem
%X Long-term personalized dialogue agents must track user preferences as their personas evolve. Existing memory systems organize past events well, but store personas as flat profiles detached from the events that justify them. This loose coupling leads to the memory–persona validity gap and the persona-aware retrieval gap. We propose PGMem, a heterogeneous persona-memory graph that connects event and persona nodes through typed provenance and evidence edges, keeping each persona signal traceable to the events that support or revise it. At retrieval time, PGMem expands from query-relevant seeds and ranks signals by evidential validity. Across three benchmarks with small language model backbones, PGMem consistently outperforms summary-based, persona-aware, graph-structured, and agentic memory baselines, and improves performance as the context grows. The source codes of PGMem are available at https://github.com/wonjunchoi23/pgmem/
%U https://aclanthology.org/2026.emnlp-main.1891/
%P 43087-43112
Markdown (Informal)
[PGMem: Tightly Coupled Persona–Memory Graph for Lifelong Personalized Agents](https://aclanthology.org/2026.emnlp-main.1891/) (Choi et al., EMNLP 2026)
ACL