@inproceedings{kim-etal-2026-talk,
title = "{T}a{LK}: Text-attributed Graph Dataset Distillation via Coupling Language Model with Graph-Aware Kernel",
author = "Kim, Yeongho and
Choi, Yeonje and
Shin, Kijung",
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.31/",
pages = "615--629",
ISBN = "979-8-89176-462-0",
abstract = "Text-attributed graphs (TAGs) are widely used in many real-world domains, and learning on TAGs requires jointly modeling text semantics and graph structure. A standard approach for modeling TAGs is to combine a language model (LM) and a graph neural network (GNN), but joint training is computationally expensive and difficult to scale. Dataset distillation is a promising way to reduce training costs, but existing methods are not well suited to TAGs because they are typically designed for a single modality or still require repeatedly training expensive LM-GNN models on the full dataset during distillation. To address this, we propose TaLK, an effective dataset distillation method for TAGs that couples an LM with a graph-aware neural tangent kernel. This design enables efficient dataset distillation, avoiding repeated joint training on the full dataset while reflecting both textual and structural information for effective TAG learning. Experiments on multiple TAG benchmarks show that TaLK consistently outperforms existing baselines and achieves up to 97{\%} of full-dataset performance with only 1{\%} synthetic data."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="kim-etal-2026-talk">
<titleInfo>
<title>TaLK: Text-attributed Graph Dataset Distillation via Coupling Language Model with Graph-Aware Kernel</title>
</titleInfo>
<name type="personal">
<namePart type="given">Yeongho</namePart>
<namePart type="family">Kim</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yeonje</namePart>
<namePart type="family">Choi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kijung</namePart>
<namePart type="family">Shin</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>Text-attributed graphs (TAGs) are widely used in many real-world domains, and learning on TAGs requires jointly modeling text semantics and graph structure. A standard approach for modeling TAGs is to combine a language model (LM) and a graph neural network (GNN), but joint training is computationally expensive and difficult to scale. Dataset distillation is a promising way to reduce training costs, but existing methods are not well suited to TAGs because they are typically designed for a single modality or still require repeatedly training expensive LM-GNN models on the full dataset during distillation. To address this, we propose TaLK, an effective dataset distillation method for TAGs that couples an LM with a graph-aware neural tangent kernel. This design enables efficient dataset distillation, avoiding repeated joint training on the full dataset while reflecting both textual and structural information for effective TAG learning. Experiments on multiple TAG benchmarks show that TaLK consistently outperforms existing baselines and achieves up to 97% of full-dataset performance with only 1% synthetic data.</abstract>
<identifier type="citekey">kim-etal-2026-talk</identifier>
<location>
<url>https://aclanthology.org/2026.findings-emnlp.31/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>615</start>
<end>629</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T TaLK: Text-attributed Graph Dataset Distillation via Coupling Language Model with Graph-Aware Kernel
%A Kim, Yeongho
%A Choi, Yeonje
%A Shin, Kijung
%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 kim-etal-2026-talk
%X Text-attributed graphs (TAGs) are widely used in many real-world domains, and learning on TAGs requires jointly modeling text semantics and graph structure. A standard approach for modeling TAGs is to combine a language model (LM) and a graph neural network (GNN), but joint training is computationally expensive and difficult to scale. Dataset distillation is a promising way to reduce training costs, but existing methods are not well suited to TAGs because they are typically designed for a single modality or still require repeatedly training expensive LM-GNN models on the full dataset during distillation. To address this, we propose TaLK, an effective dataset distillation method for TAGs that couples an LM with a graph-aware neural tangent kernel. This design enables efficient dataset distillation, avoiding repeated joint training on the full dataset while reflecting both textual and structural information for effective TAG learning. Experiments on multiple TAG benchmarks show that TaLK consistently outperforms existing baselines and achieves up to 97% of full-dataset performance with only 1% synthetic data.
%U https://aclanthology.org/2026.findings-emnlp.31/
%P 615-629
Markdown (Informal)
[TaLK: Text-attributed Graph Dataset Distillation via Coupling Language Model with Graph-Aware Kernel](https://aclanthology.org/2026.findings-emnlp.31/) (Kim et al., Findings 2026)
ACL