Ego-based Entropy Measures for Structural Representations on Graphs

Machine learning on graph-structured data has attracted high research interest due to the emergence of Graph Neural Networks (GNNs). Most of the proposed GNNs are based on the node homophily, i.e neighboring nodes share similar characteristics. However, in many complex networks, nodes that lie to di...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Dasoulas, George, Nikolentzos, Giannis, Scaman, Kevin, Virmaux, Aladin, Vazirgiannis, Michalis
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page
container_issue
container_start_page
container_title
container_volume
creator Dasoulas, George
Nikolentzos, Giannis
Scaman, Kevin
Virmaux, Aladin
Vazirgiannis, Michalis
description Machine learning on graph-structured data has attracted high research interest due to the emergence of Graph Neural Networks (GNNs). Most of the proposed GNNs are based on the node homophily, i.e neighboring nodes share similar characteristics. However, in many complex networks, nodes that lie to distant parts of the graph share structurally equivalent characteristics and exhibit similar roles (e.g chemical properties of distant atoms in a molecule, type of social network users). A growing literature proposed representations that identify structurally equivalent nodes. However, most of the existing methods require high time and space complexity. In this paper, we propose VNEstruct, a simple approach, based on entropy measures of the neighborhood's topology, for generating low-dimensional structural representations, that is time-efficient and robust to graph perturbations. Empirically, we observe that VNEstruct exhibits robustness on structural role identification tasks. Moreover, VNEstruct can achieve state-of-the-art performance on graph classification, without incorporating the graph structure information in the optimization, in contrast to GNN competitors.
doi_str_mv 10.48550/arxiv.2102.08735
format Article
fullrecord <record><control><sourceid>arxiv_GOX</sourceid><recordid>TN_cdi_arxiv_primary_2102_08735</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2102_08735</sourcerecordid><originalsourceid>FETCH-LOGICAL-a675-b7151a58c054be883ad2d71b4b63c5da806166fd4a5bb336c17fd696024993f3</originalsourceid><addsrcrecordid>eNotz81KxDAYheFsXMjoBbgyN9CaNL9dSqmjMDLguC9fmkQLY1O-pOLcvTq6OvAuDjyE3HBWS6sUuwP8mj7rhrOmZtYIdUm6_i1VDnLwtJ8LpuVEnwPkFUOmMSE9FFzHsiIc6UtYfmqYC5QpzZmmmW4Rlvd8RS4iHHO4_t8NOTz0r91jtdtvn7r7XQXaqMoZrjgoOzIlXbBWgG-84U46LUblwTLNtY5egnJOCD1yE71uNWtk24ooNuT27_WMGBacPgBPwy9mOGPEN9-eRJQ</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>Ego-based Entropy Measures for Structural Representations on Graphs</title><source>arXiv.org</source><creator>Dasoulas, George ; Nikolentzos, Giannis ; Scaman, Kevin ; Virmaux, Aladin ; Vazirgiannis, Michalis</creator><creatorcontrib>Dasoulas, George ; Nikolentzos, Giannis ; Scaman, Kevin ; Virmaux, Aladin ; Vazirgiannis, Michalis</creatorcontrib><description>Machine learning on graph-structured data has attracted high research interest due to the emergence of Graph Neural Networks (GNNs). Most of the proposed GNNs are based on the node homophily, i.e neighboring nodes share similar characteristics. However, in many complex networks, nodes that lie to distant parts of the graph share structurally equivalent characteristics and exhibit similar roles (e.g chemical properties of distant atoms in a molecule, type of social network users). A growing literature proposed representations that identify structurally equivalent nodes. However, most of the existing methods require high time and space complexity. In this paper, we propose VNEstruct, a simple approach, based on entropy measures of the neighborhood's topology, for generating low-dimensional structural representations, that is time-efficient and robust to graph perturbations. Empirically, we observe that VNEstruct exhibits robustness on structural role identification tasks. Moreover, VNEstruct can achieve state-of-the-art performance on graph classification, without incorporating the graph structure information in the optimization, in contrast to GNN competitors.</description><identifier>DOI: 10.48550/arxiv.2102.08735</identifier><language>eng</language><subject>Computer Science - Learning ; Computer Science - Social and Information Networks</subject><creationdate>2021-02</creationdate><rights>http://arxiv.org/licenses/nonexclusive-distrib/1.0</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>228,230,776,881</link.rule.ids><linktorsrc>$$Uhttps://arxiv.org/abs/2102.08735$$EView_record_in_Cornell_University$$FView_record_in_$$GCornell_University$$Hfree_for_read</linktorsrc><backlink>$$Uhttps://doi.org/10.48550/arXiv.2102.08735$$DView paper in arXiv$$Hfree_for_read</backlink></links><search><creatorcontrib>Dasoulas, George</creatorcontrib><creatorcontrib>Nikolentzos, Giannis</creatorcontrib><creatorcontrib>Scaman, Kevin</creatorcontrib><creatorcontrib>Virmaux, Aladin</creatorcontrib><creatorcontrib>Vazirgiannis, Michalis</creatorcontrib><title>Ego-based Entropy Measures for Structural Representations on Graphs</title><description>Machine learning on graph-structured data has attracted high research interest due to the emergence of Graph Neural Networks (GNNs). Most of the proposed GNNs are based on the node homophily, i.e neighboring nodes share similar characteristics. However, in many complex networks, nodes that lie to distant parts of the graph share structurally equivalent characteristics and exhibit similar roles (e.g chemical properties of distant atoms in a molecule, type of social network users). A growing literature proposed representations that identify structurally equivalent nodes. However, most of the existing methods require high time and space complexity. In this paper, we propose VNEstruct, a simple approach, based on entropy measures of the neighborhood's topology, for generating low-dimensional structural representations, that is time-efficient and robust to graph perturbations. Empirically, we observe that VNEstruct exhibits robustness on structural role identification tasks. Moreover, VNEstruct can achieve state-of-the-art performance on graph classification, without incorporating the graph structure information in the optimization, in contrast to GNN competitors.</description><subject>Computer Science - Learning</subject><subject>Computer Science - Social and Information Networks</subject><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2021</creationdate><recordtype>article</recordtype><sourceid>GOX</sourceid><recordid>eNotz81KxDAYheFsXMjoBbgyN9CaNL9dSqmjMDLguC9fmkQLY1O-pOLcvTq6OvAuDjyE3HBWS6sUuwP8mj7rhrOmZtYIdUm6_i1VDnLwtJ8LpuVEnwPkFUOmMSE9FFzHsiIc6UtYfmqYC5QpzZmmmW4Rlvd8RS4iHHO4_t8NOTz0r91jtdtvn7r7XQXaqMoZrjgoOzIlXbBWgG-84U46LUblwTLNtY5egnJOCD1yE71uNWtk24ooNuT27_WMGBacPgBPwy9mOGPEN9-eRJQ</recordid><startdate>20210217</startdate><enddate>20210217</enddate><creator>Dasoulas, George</creator><creator>Nikolentzos, Giannis</creator><creator>Scaman, Kevin</creator><creator>Virmaux, Aladin</creator><creator>Vazirgiannis, Michalis</creator><scope>AKY</scope><scope>GOX</scope></search><sort><creationdate>20210217</creationdate><title>Ego-based Entropy Measures for Structural Representations on Graphs</title><author>Dasoulas, George ; Nikolentzos, Giannis ; Scaman, Kevin ; Virmaux, Aladin ; Vazirgiannis, Michalis</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-a675-b7151a58c054be883ad2d71b4b63c5da806166fd4a5bb336c17fd696024993f3</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2021</creationdate><topic>Computer Science - Learning</topic><topic>Computer Science - Social and Information Networks</topic><toplevel>online_resources</toplevel><creatorcontrib>Dasoulas, George</creatorcontrib><creatorcontrib>Nikolentzos, Giannis</creatorcontrib><creatorcontrib>Scaman, Kevin</creatorcontrib><creatorcontrib>Virmaux, Aladin</creatorcontrib><creatorcontrib>Vazirgiannis, Michalis</creatorcontrib><collection>arXiv Computer Science</collection><collection>arXiv.org</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Dasoulas, George</au><au>Nikolentzos, Giannis</au><au>Scaman, Kevin</au><au>Virmaux, Aladin</au><au>Vazirgiannis, Michalis</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Ego-based Entropy Measures for Structural Representations on Graphs</atitle><date>2021-02-17</date><risdate>2021</risdate><abstract>Machine learning on graph-structured data has attracted high research interest due to the emergence of Graph Neural Networks (GNNs). Most of the proposed GNNs are based on the node homophily, i.e neighboring nodes share similar characteristics. However, in many complex networks, nodes that lie to distant parts of the graph share structurally equivalent characteristics and exhibit similar roles (e.g chemical properties of distant atoms in a molecule, type of social network users). A growing literature proposed representations that identify structurally equivalent nodes. However, most of the existing methods require high time and space complexity. In this paper, we propose VNEstruct, a simple approach, based on entropy measures of the neighborhood's topology, for generating low-dimensional structural representations, that is time-efficient and robust to graph perturbations. Empirically, we observe that VNEstruct exhibits robustness on structural role identification tasks. Moreover, VNEstruct can achieve state-of-the-art performance on graph classification, without incorporating the graph structure information in the optimization, in contrast to GNN competitors.</abstract><doi>10.48550/arxiv.2102.08735</doi><oa>free_for_read</oa></addata></record>
fulltext fulltext_linktorsrc
identifier DOI: 10.48550/arxiv.2102.08735
ispartof
issn
language eng
recordid cdi_arxiv_primary_2102_08735
source arXiv.org
subjects Computer Science - Learning
Computer Science - Social and Information Networks
title Ego-based Entropy Measures for Structural Representations on Graphs
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-29T05%3A33%3A03IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-arxiv_GOX&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Ego-based%20Entropy%20Measures%20for%20Structural%20Representations%20on%20Graphs&rft.au=Dasoulas,%20George&rft.date=2021-02-17&rft_id=info:doi/10.48550/arxiv.2102.08735&rft_dat=%3Carxiv_GOX%3E2102_08735%3C/arxiv_GOX%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rfr_iscdi=true