Efficient batched distance and centrality computation in unweighted and weighted graphs

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Scopus citations

Abstract

Distance and centrality computations are important building blocks for modern graph databases as well as for dedicated graph analytics systems. Two commonly used centrality metrics are the compute-intense closeness and betweenness centralities, which require numerous expensive shortest distance calculations. We propose batched algorithm execution to run multiple distance and centrality computations at the same time and let them share common graph and data accesses. Batched execution amortizes the high cost of random memory accesses and presents new vectorization potential on modern CPUs and compute accelerators. We show how batched algorithm execution can be leveraged to significantly improve the performance of distance, closeness, and betweenness centrality calculations on unweighted and weighted graphs. Our evaluation demonstrates that batched execution can improve the runtime of these common metrics by over an order of magnitude.

Original languageEnglish
Title of host publicationDatenbanksysteme fur Business, Technologie und Web, BTW 2017 - 17. Fachtagung des GI-Fachbereichs "Datenbanken und Informationssysteme�, DBIS 2017, Proceedings
EditorsBernhard Mitschang, Daniela Nicklas, Frank Leymann, Harald Schoning, Melanie Herschel, Jens Teubner, Theo Harder
PublisherGesellschaft fur Informatik (GI)
Pages247-266
Number of pages20
ISBN (Electronic)9783885796596
StatePublished - 2017
EventDatenbanksysteme fur Business, Technologie und Web, BTW 2017, 17. Fachtagung des GI-Fachbereichs "Datenbanken und Informationssysteme�, DBIS - Database Systems for Business, Technology and Web, BTW 2017, 17th Symposium of the GI Department "Databases and Information Systems", DBIS - Stuttgart, Germany
Duration: 6 Mar 201710 Mar 2017

Publication series

NameLecture Notes in Informatics (LNI), Proceedings - Series of the Gesellschaft fur Informatik (GI)
Volume265
ISSN (Print)1617-5468

Conference

ConferenceDatenbanksysteme fur Business, Technologie und Web, BTW 2017, 17. Fachtagung des GI-Fachbereichs "Datenbanken und Informationssysteme�, DBIS - Database Systems for Business, Technology and Web, BTW 2017, 17th Symposium of the GI Department "Databases and Information Systems", DBIS
Country/TerritoryGermany
CityStuttgart
Period6/03/1710/03/17

Keywords

  • Betweenness centrality
  • Closeness centrality
  • Graph analytics
  • Graph databases

Fingerprint

Dive into the research topics of 'Efficient batched distance and centrality computation in unweighted and weighted graphs'. Together they form a unique fingerprint.

Cite this