Cargando…

ABCDE: Approximating Betweenness-Centrality ranking with progressive-DropEdge

Betweenness-centrality is a popular measure in network analysis that aims to describe the importance of nodes in a graph. It accounts for the fraction of shortest paths passing through that node and is a key measure in many applications including community detection and network dismantling. The comp...

Descripción completa

Detalles Bibliográficos
Autor principal: Mirakyan, Martin
Formato: Online Artículo Texto
Lenguaje:English
Publicado: PeerJ Inc. 2021
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8444073/
https://www.ncbi.nlm.nih.gov/pubmed/34604524
http://dx.doi.org/10.7717/peerj-cs.699
_version_ 1784568416682639360
author Mirakyan, Martin
author_facet Mirakyan, Martin
author_sort Mirakyan, Martin
collection PubMed
description Betweenness-centrality is a popular measure in network analysis that aims to describe the importance of nodes in a graph. It accounts for the fraction of shortest paths passing through that node and is a key measure in many applications including community detection and network dismantling. The computation of betweenness-centrality for each node in a graph requires an excessive amount of computing power, especially for large graphs. On the other hand, in many applications, the main interest lies in finding the top-k most important nodes in the graph. Therefore, several approximation algorithms were proposed to solve the problem faster. Some recent approaches propose to use shallow graph convolutional networks to approximate the top-k nodes with the highest betweenness-centrality scores. This work presents a deep graph convolutional neural network that outputs a rank score for each node in a given graph. With careful optimization and regularization tricks, including an extended version of DropEdge which is named Progressive-DropEdge, the system achieves better results than the current approaches. Experiments on both real-world and synthetic datasets show that the presented algorithm is an order of magnitude faster in inference and requires several times fewer resources and time to train.
format Online
Article
Text
id pubmed-8444073
institution National Center for Biotechnology Information
language English
publishDate 2021
publisher PeerJ Inc.
record_format MEDLINE/PubMed
spelling pubmed-84440732021-09-30 ABCDE: Approximating Betweenness-Centrality ranking with progressive-DropEdge Mirakyan, Martin PeerJ Comput Sci Algorithms and Analysis of Algorithms Betweenness-centrality is a popular measure in network analysis that aims to describe the importance of nodes in a graph. It accounts for the fraction of shortest paths passing through that node and is a key measure in many applications including community detection and network dismantling. The computation of betweenness-centrality for each node in a graph requires an excessive amount of computing power, especially for large graphs. On the other hand, in many applications, the main interest lies in finding the top-k most important nodes in the graph. Therefore, several approximation algorithms were proposed to solve the problem faster. Some recent approaches propose to use shallow graph convolutional networks to approximate the top-k nodes with the highest betweenness-centrality scores. This work presents a deep graph convolutional neural network that outputs a rank score for each node in a given graph. With careful optimization and regularization tricks, including an extended version of DropEdge which is named Progressive-DropEdge, the system achieves better results than the current approaches. Experiments on both real-world and synthetic datasets show that the presented algorithm is an order of magnitude faster in inference and requires several times fewer resources and time to train. PeerJ Inc. 2021-09-06 /pmc/articles/PMC8444073/ /pubmed/34604524 http://dx.doi.org/10.7717/peerj-cs.699 Text en © 2021 Mirakyan https://creativecommons.org/licenses/by/4.0/This is an open access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/) , which permits unrestricted use, distribution, reproduction and adaptation in any medium and for any purpose provided that it is properly attributed. For attribution, the original author(s), title, publication source (PeerJ Computer Science) and either DOI or URL of the article must be cited.
spellingShingle Algorithms and Analysis of Algorithms
Mirakyan, Martin
ABCDE: Approximating Betweenness-Centrality ranking with progressive-DropEdge
title ABCDE: Approximating Betweenness-Centrality ranking with progressive-DropEdge
title_full ABCDE: Approximating Betweenness-Centrality ranking with progressive-DropEdge
title_fullStr ABCDE: Approximating Betweenness-Centrality ranking with progressive-DropEdge
title_full_unstemmed ABCDE: Approximating Betweenness-Centrality ranking with progressive-DropEdge
title_short ABCDE: Approximating Betweenness-Centrality ranking with progressive-DropEdge
title_sort abcde: approximating betweenness-centrality ranking with progressive-dropedge
topic Algorithms and Analysis of Algorithms
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8444073/
https://www.ncbi.nlm.nih.gov/pubmed/34604524
http://dx.doi.org/10.7717/peerj-cs.699
work_keys_str_mv AT mirakyanmartin abcdeapproximatingbetweennesscentralityrankingwithprogressivedropedge