@inproceedings{alikasifoglu-etal-2024-vispool,
title = "{VISP}ool: Enhancing Transformer Encoders with Vector Visibility Graph Neural Networks",
author = "Alika{\c{s}}ifo{\u{g}}lu, Tuna and
Aras, Arda and
Koc, Aykut",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Findings of the Association for Computational Linguistics ACL 2024",
month = aug,
year = "2024",
address = "Bangkok, Thailand and virtual meeting",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.findings-acl.149",
pages = "2547--2556",
abstract = "The emergence of transformers has revolutionized natural language processing (NLP), as evidenced in various NLP tasks. While graph neural networks (GNNs) show recent promise in NLP, they are not standalone replacements for transformers. Rather, recent research explores combining transformers and GNNs. Existing GNN-based approaches rely on static graph construction methods requiring excessive text processing, and most of them are not scalable with the increasing document and word counts. We address these limitations by proposing a novel dynamic graph construction method for text documents based on vector visibility graphs (VVGs) generated from transformer output. Then, we introduce visibility pooler (VISPool), a scalable model architecture that seamlessly integrates VVG convolutional networks into transformer pipelines. We evaluate the proposed model on the General Language Understanding Evaluation (GLUE) benchmark datasets. VISPool outperforms the baselines with less trainable parameters, demonstrating the viability of the visibility-based graph construction method for enhancing transformers with GNNs.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="alikasifoglu-etal-2024-vispool">
<titleInfo>
<title>VISPool: Enhancing Transformer Encoders with Vector Visibility Graph Neural Networks</title>
</titleInfo>
<name type="personal">
<namePart type="given">Tuna</namePart>
<namePart type="family">Alikaşifoğlu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Arda</namePart>
<namePart type="family">Aras</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Aykut</namePart>
<namePart type="family">Koc</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>Findings of the Association for Computational Linguistics ACL 2024</title>
</titleInfo>
<name type="personal">
<namePart type="given">Lun-Wei</namePart>
<namePart type="family">Ku</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Andre</namePart>
<namePart type="family">Martins</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Vivek</namePart>
<namePart type="family">Srikumar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Bangkok, Thailand and virtual meeting</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>The emergence of transformers has revolutionized natural language processing (NLP), as evidenced in various NLP tasks. While graph neural networks (GNNs) show recent promise in NLP, they are not standalone replacements for transformers. Rather, recent research explores combining transformers and GNNs. Existing GNN-based approaches rely on static graph construction methods requiring excessive text processing, and most of them are not scalable with the increasing document and word counts. We address these limitations by proposing a novel dynamic graph construction method for text documents based on vector visibility graphs (VVGs) generated from transformer output. Then, we introduce visibility pooler (VISPool), a scalable model architecture that seamlessly integrates VVG convolutional networks into transformer pipelines. We evaluate the proposed model on the General Language Understanding Evaluation (GLUE) benchmark datasets. VISPool outperforms the baselines with less trainable parameters, demonstrating the viability of the visibility-based graph construction method for enhancing transformers with GNNs.</abstract>
<identifier type="citekey">alikasifoglu-etal-2024-vispool</identifier>
<location>
<url>https://aclanthology.org/2024.findings-acl.149</url>
</location>
<part>
<date>2024-08</date>
<extent unit="page">
<start>2547</start>
<end>2556</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T VISPool: Enhancing Transformer Encoders with Vector Visibility Graph Neural Networks
%A Alikaşifoğlu, Tuna
%A Aras, Arda
%A Koc, Aykut
%Y Ku, Lun-Wei
%Y Martins, Andre
%Y Srikumar, Vivek
%S Findings of the Association for Computational Linguistics ACL 2024
%D 2024
%8 August
%I Association for Computational Linguistics
%C Bangkok, Thailand and virtual meeting
%F alikasifoglu-etal-2024-vispool
%X The emergence of transformers has revolutionized natural language processing (NLP), as evidenced in various NLP tasks. While graph neural networks (GNNs) show recent promise in NLP, they are not standalone replacements for transformers. Rather, recent research explores combining transformers and GNNs. Existing GNN-based approaches rely on static graph construction methods requiring excessive text processing, and most of them are not scalable with the increasing document and word counts. We address these limitations by proposing a novel dynamic graph construction method for text documents based on vector visibility graphs (VVGs) generated from transformer output. Then, we introduce visibility pooler (VISPool), a scalable model architecture that seamlessly integrates VVG convolutional networks into transformer pipelines. We evaluate the proposed model on the General Language Understanding Evaluation (GLUE) benchmark datasets. VISPool outperforms the baselines with less trainable parameters, demonstrating the viability of the visibility-based graph construction method for enhancing transformers with GNNs.
%U https://aclanthology.org/2024.findings-acl.149
%P 2547-2556
Markdown (Informal)
[VISPool: Enhancing Transformer Encoders with Vector Visibility Graph Neural Networks](https://aclanthology.org/2024.findings-acl.149) (Alikaşifoğlu et al., Findings 2024)
ACL