Beyond Weisfeiler–Lehman with Local Ego-Network Encodings

Identifying similar network structures is key to capturing graph isomorphisms and learning representations that exploit structural information encoded in graph data. This work shows that ego networks can produce a structural encoding scheme for arbitrary graphs with greater expressivity than the Wei...

Descripción completa

Detalles Bibliográficos
Autores: Alvarez-Gonzalez, Nurudin, Kaltenbrunner, Andreas, Gómez, Vicenç
Tipo de recurso: artículo
Estado:Versión publicada
Fecha de publicación:2023
País:España
Institución:Universitat Oberta de Catalunya (UOC)
Repositorio:O2, repositorio institucional de la UOC
OAI Identifier:oai:openaccess.uoc.edu:10609/150374
Acceso en línea:http://hdl.handle.net/10609/150374
https://doi.org/10.3390/make5040063
Access Level:acceso abierto
Palabra clave:graph neural networks
graph representation learning
weisfeiler–lehman
graph isomorphism
GNN expressivity
ego networks
id ES_81fa2b054cd7e0f4c0fa82dc1a92c80d
oai_identifier_str oai:openaccess.uoc.edu:10609/150374
network_acronym_str ES
network_name_str España
repository_id_str
spelling Beyond Weisfeiler–Lehman with Local Ego-Network EncodingsAlvarez-Gonzalez, NurudinKaltenbrunner, AndreasGómez, Vicençgraph neural networksgraph representation learningweisfeiler–lehmangraph isomorphismGNN expressivityego networksIdentifying similar network structures is key to capturing graph isomorphisms and learning representations that exploit structural information encoded in graph data. This work shows that ego networks can produce a structural encoding scheme for arbitrary graphs with greater expressivity than the Weisfeiler–Lehman (1-WL) test. We introduce IGEL, a preprocessing step to produce features that augment node representations by encoding ego networks into sparse vectors that enrich message passing (MP) graph neural networks (GNNs) beyond 1-WL expressivity. We formally describe the relation between IGEL and 1-WL, and characterize its expressive power and limitations. Experiments show that IGEL matches the empirical expressivity of state-of-the-art methods on isomorphism detection while improving performance on nine GNN architectures and six graph machine learning tasks.MDPI AG202420242023info:eu-repo/semantics/articleinfo:eu-repo/semantics/publishedVersionapplication/pdfapplication/pdfhttp://hdl.handle.net/10609/150374https://doi.org/10.3390/make5040063reponame:O2, repositorio institucional de la UOCinstname:Universitat Oberta de Catalunya (UOC)InglésMachine Learning and Knowledge Extraction (MAKE), 2023, 5(4)https://doi.org/10.3390/make5040063info:eu-repo/grantAgreement/MCIN/AEI/CEX2021-001195-M ; info:eu-repoCC BYhttps://creativecommons.org/licenses/by/4.0/info:eu-repo/semantics/openAccessoai:openaccess.uoc.edu:10609/1503742026-05-28T12:42:01Z
dc.title.none.fl_str_mv Beyond Weisfeiler–Lehman with Local Ego-Network Encodings
title Beyond Weisfeiler–Lehman with Local Ego-Network Encodings
spellingShingle Beyond Weisfeiler–Lehman with Local Ego-Network Encodings
Alvarez-Gonzalez, Nurudin
graph neural networks
graph representation learning
weisfeiler–lehman
graph isomorphism
GNN expressivity
ego networks
title_short Beyond Weisfeiler–Lehman with Local Ego-Network Encodings
title_full Beyond Weisfeiler–Lehman with Local Ego-Network Encodings
title_fullStr Beyond Weisfeiler–Lehman with Local Ego-Network Encodings
title_full_unstemmed Beyond Weisfeiler–Lehman with Local Ego-Network Encodings
title_sort Beyond Weisfeiler–Lehman with Local Ego-Network Encodings
dc.creator.none.fl_str_mv Alvarez-Gonzalez, Nurudin
Kaltenbrunner, Andreas
Gómez, Vicenç
author Alvarez-Gonzalez, Nurudin
author_facet Alvarez-Gonzalez, Nurudin
Kaltenbrunner, Andreas
Gómez, Vicenç
author_role author
author2 Kaltenbrunner, Andreas
Gómez, Vicenç
author2_role author
author
dc.subject.none.fl_str_mv graph neural networks
graph representation learning
weisfeiler–lehman
graph isomorphism
GNN expressivity
ego networks
topic graph neural networks
graph representation learning
weisfeiler–lehman
graph isomorphism
GNN expressivity
ego networks
description Identifying similar network structures is key to capturing graph isomorphisms and learning representations that exploit structural information encoded in graph data. This work shows that ego networks can produce a structural encoding scheme for arbitrary graphs with greater expressivity than the Weisfeiler–Lehman (1-WL) test. We introduce IGEL, a preprocessing step to produce features that augment node representations by encoding ego networks into sparse vectors that enrich message passing (MP) graph neural networks (GNNs) beyond 1-WL expressivity. We formally describe the relation between IGEL and 1-WL, and characterize its expressive power and limitations. Experiments show that IGEL matches the empirical expressivity of state-of-the-art methods on isomorphism detection while improving performance on nine GNN architectures and six graph machine learning tasks.
publishDate 2023
dc.date.none.fl_str_mv 2023
2024
2024
dc.type.none.fl_str_mv info:eu-repo/semantics/article
info:eu-repo/semantics/publishedVersion
format article
status_str publishedVersion
dc.identifier.none.fl_str_mv http://hdl.handle.net/10609/150374
https://doi.org/10.3390/make5040063
url http://hdl.handle.net/10609/150374
https://doi.org/10.3390/make5040063
dc.language.none.fl_str_mv Inglés
language_invalid_str_mv Inglés
dc.relation.none.fl_str_mv Machine Learning and Knowledge Extraction (MAKE), 2023, 5(4)
https://doi.org/10.3390/make5040063
info:eu-repo/grantAgreement/MCIN/AEI/CEX2021-001195-M ; info:eu-repo
dc.rights.none.fl_str_mv CC BY
https://creativecommons.org/licenses/by/4.0/
info:eu-repo/semantics/openAccess
rights_invalid_str_mv CC BY
https://creativecommons.org/licenses/by/4.0/
eu_rights_str_mv openAccess
dc.format.none.fl_str_mv application/pdf
application/pdf
dc.publisher.none.fl_str_mv MDPI AG
publisher.none.fl_str_mv MDPI AG
dc.source.none.fl_str_mv reponame:O2, repositorio institucional de la UOC
instname:Universitat Oberta de Catalunya (UOC)
instname_str Universitat Oberta de Catalunya (UOC)
reponame_str O2, repositorio institucional de la UOC
collection O2, repositorio institucional de la UOC
repository.name.fl_str_mv
repository.mail.fl_str_mv
_version_ 1869412013189365760
score 15,81155

Similares en LA Referencia