@inproceedings{zolnai-lucas-etal-2024-stage,
title = "{STAGE}: Simplified Text-Attributed Graph Embeddings using Pre-trained {LLM}s",
author = "Zolnai-Lucas, Aaron and
Boylan, Jack and
Hokamp, Chris and
Ghaffari, Parsa",
editor = "Biswas, Russa and
Kaffee, Lucie-Aim{\'e}e and
Agarwal, Oshin and
Minervini, Pasquale and
Singh, Sameer and
de Melo, Gerard",
booktitle = "Proceedings of the 1st Workshop on Knowledge Graphs and Large Language Models (KaLLM 2024)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.kallm-1.10",
doi = "10.18653/v1/2024.kallm-1.10",
pages = "92--104",
abstract = "We present STAGE, a straightforward yet effective method for enhancing node features in Graph Neural Network (GNN) models that encode Text-Attributed Graphs (TAGs). Our approach leverages Large-Language Models (LLMs) to generate embeddings for textual attributes. STAGE achieves competitive results on various node classification benchmarks while also maintaining a simplicity in implementation relative to current state-of-the-art (SoTA) techniques. We show that utilizing pre-trained LLMs as embedding generators provides robust features for ensemble GNN training, enabling pipelines that are simpler than current SoTA approaches which require multiple expensive training and prompting stages. We also implement diffusion-pattern GNNs in an effort to make this pipeline scalable to graphs beyond academic benchmarks.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="zolnai-lucas-etal-2024-stage">
<titleInfo>
<title>STAGE: Simplified Text-Attributed Graph Embeddings using Pre-trained LLMs</title>
</titleInfo>
<name type="personal">
<namePart type="given">Aaron</namePart>
<namePart type="family">Zolnai-Lucas</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jack</namePart>
<namePart type="family">Boylan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chris</namePart>
<namePart type="family">Hokamp</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Parsa</namePart>
<namePart type="family">Ghaffari</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2024-08</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 1st Workshop on Knowledge Graphs and Large Language Models (KaLLM 2024)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Russa</namePart>
<namePart type="family">Biswas</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Lucie-Aimée</namePart>
<namePart type="family">Kaffee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Oshin</namePart>
<namePart type="family">Agarwal</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Pasquale</namePart>
<namePart type="family">Minervini</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sameer</namePart>
<namePart type="family">Singh</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Gerard</namePart>
<namePart type="family">de Melo</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Bangkok, Thailand</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>We present STAGE, a straightforward yet effective method for enhancing node features in Graph Neural Network (GNN) models that encode Text-Attributed Graphs (TAGs). Our approach leverages Large-Language Models (LLMs) to generate embeddings for textual attributes. STAGE achieves competitive results on various node classification benchmarks while also maintaining a simplicity in implementation relative to current state-of-the-art (SoTA) techniques. We show that utilizing pre-trained LLMs as embedding generators provides robust features for ensemble GNN training, enabling pipelines that are simpler than current SoTA approaches which require multiple expensive training and prompting stages. We also implement diffusion-pattern GNNs in an effort to make this pipeline scalable to graphs beyond academic benchmarks.</abstract>
<identifier type="citekey">zolnai-lucas-etal-2024-stage</identifier>
<identifier type="doi">10.18653/v1/2024.kallm-1.10</identifier>
<location>
<url>https://aclanthology.org/2024.kallm-1.10</url>
</location>
<part>
<date>2024-08</date>
<extent unit="page">
<start>92</start>
<end>104</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T STAGE: Simplified Text-Attributed Graph Embeddings using Pre-trained LLMs
%A Zolnai-Lucas, Aaron
%A Boylan, Jack
%A Hokamp, Chris
%A Ghaffari, Parsa
%Y Biswas, Russa
%Y Kaffee, Lucie-Aimée
%Y Agarwal, Oshin
%Y Minervini, Pasquale
%Y Singh, Sameer
%Y de Melo, Gerard
%S Proceedings of the 1st Workshop on Knowledge Graphs and Large Language Models (KaLLM 2024)
%D 2024
%8 August
%I Association for Computational Linguistics
%C Bangkok, Thailand
%F zolnai-lucas-etal-2024-stage
%X We present STAGE, a straightforward yet effective method for enhancing node features in Graph Neural Network (GNN) models that encode Text-Attributed Graphs (TAGs). Our approach leverages Large-Language Models (LLMs) to generate embeddings for textual attributes. STAGE achieves competitive results on various node classification benchmarks while also maintaining a simplicity in implementation relative to current state-of-the-art (SoTA) techniques. We show that utilizing pre-trained LLMs as embedding generators provides robust features for ensemble GNN training, enabling pipelines that are simpler than current SoTA approaches which require multiple expensive training and prompting stages. We also implement diffusion-pattern GNNs in an effort to make this pipeline scalable to graphs beyond academic benchmarks.
%R 10.18653/v1/2024.kallm-1.10
%U https://aclanthology.org/2024.kallm-1.10
%U https://doi.org/10.18653/v1/2024.kallm-1.10
%P 92-104
Markdown (Informal)
[STAGE: Simplified Text-Attributed Graph Embeddings using Pre-trained LLMs](https://aclanthology.org/2024.kallm-1.10) (Zolnai-Lucas et al., KaLLM-WS 2024)
ACL