Parallel k-Core decomposition on multicore platforms

Humayun Kabir, Kamesh Madduri

Research output: Chapter in Book/Report/Conference proceedingConference contribution

9 Citations (Scopus)

Abstract

k-core decomposition is a social network analytic that can be applied to centrality analysis, visualization, and community detection. There is a simple and well-known linear time sequential algorithm to perform k-core decomposition, and for this reason, k-core decomposition is widely used as a network analytic. In this work, we present a new shared-memory parallel algorithm called PKC for k-core decomposition on multicore platforms. This approach improves on the state-of-the-art implementations for k-core decomposition algorithms by reducing synchronization overhead and creating a smaller graph to process high-degree vertices. We show that PKC consistently outperforms implementations of other methods on a 32-core multicore server and on a collection of large sparse graphs. We achieve a 2.81× speedup (geometric mean of speedups for 17 large graphs) over our implementation of the ParK k-core decomposition algorithm.

Original languageEnglish (US)
Title of host publicationProceedings - 2017 IEEE 31st International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2017
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages1482-1491
Number of pages10
ISBN (Electronic)9781538634080
DOIs
StatePublished - Jun 30 2017
Event31st IEEE International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2017 - Orlando, United States
Duration: May 29 2017Jun 2 2017

Publication series

NameProceedings - 2017 IEEE 31st International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2017

Other

Other31st IEEE International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2017
CountryUnited States
CityOrlando
Period5/29/176/2/17

Fingerprint

Decomposition
Parallel algorithms
Synchronization
Servers
Visualization
Data storage equipment

All Science Journal Classification (ASJC) codes

  • Hardware and Architecture
  • Computer Networks and Communications
  • Information Systems

Cite this

Kabir, H., & Madduri, K. (2017). Parallel k-Core decomposition on multicore platforms. In Proceedings - 2017 IEEE 31st International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2017 (pp. 1482-1491). [7965211] (Proceedings - 2017 IEEE 31st International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2017). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/IPDPSW.2017.151
Kabir, Humayun ; Madduri, Kamesh. / Parallel k-Core decomposition on multicore platforms. Proceedings - 2017 IEEE 31st International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2017. Institute of Electrical and Electronics Engineers Inc., 2017. pp. 1482-1491 (Proceedings - 2017 IEEE 31st International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2017).
@inproceedings{80f4050b06ab4a598f0f58def32e7583,
title = "Parallel k-Core decomposition on multicore platforms",
abstract = "k-core decomposition is a social network analytic that can be applied to centrality analysis, visualization, and community detection. There is a simple and well-known linear time sequential algorithm to perform k-core decomposition, and for this reason, k-core decomposition is widely used as a network analytic. In this work, we present a new shared-memory parallel algorithm called PKC for k-core decomposition on multicore platforms. This approach improves on the state-of-the-art implementations for k-core decomposition algorithms by reducing synchronization overhead and creating a smaller graph to process high-degree vertices. We show that PKC consistently outperforms implementations of other methods on a 32-core multicore server and on a collection of large sparse graphs. We achieve a 2.81× speedup (geometric mean of speedups for 17 large graphs) over our implementation of the ParK k-core decomposition algorithm.",
author = "Humayun Kabir and Kamesh Madduri",
year = "2017",
month = "6",
day = "30",
doi = "10.1109/IPDPSW.2017.151",
language = "English (US)",
series = "Proceedings - 2017 IEEE 31st International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2017",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
pages = "1482--1491",
booktitle = "Proceedings - 2017 IEEE 31st International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2017",
address = "United States",

}

Kabir, H & Madduri, K 2017, Parallel k-Core decomposition on multicore platforms. in Proceedings - 2017 IEEE 31st International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2017., 7965211, Proceedings - 2017 IEEE 31st International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2017, Institute of Electrical and Electronics Engineers Inc., pp. 1482-1491, 31st IEEE International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2017, Orlando, United States, 5/29/17. https://doi.org/10.1109/IPDPSW.2017.151

Parallel k-Core decomposition on multicore platforms. / Kabir, Humayun; Madduri, Kamesh.

Proceedings - 2017 IEEE 31st International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2017. Institute of Electrical and Electronics Engineers Inc., 2017. p. 1482-1491 7965211 (Proceedings - 2017 IEEE 31st International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2017).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Parallel k-Core decomposition on multicore platforms

AU - Kabir, Humayun

AU - Madduri, Kamesh

PY - 2017/6/30

Y1 - 2017/6/30

N2 - k-core decomposition is a social network analytic that can be applied to centrality analysis, visualization, and community detection. There is a simple and well-known linear time sequential algorithm to perform k-core decomposition, and for this reason, k-core decomposition is widely used as a network analytic. In this work, we present a new shared-memory parallel algorithm called PKC for k-core decomposition on multicore platforms. This approach improves on the state-of-the-art implementations for k-core decomposition algorithms by reducing synchronization overhead and creating a smaller graph to process high-degree vertices. We show that PKC consistently outperforms implementations of other methods on a 32-core multicore server and on a collection of large sparse graphs. We achieve a 2.81× speedup (geometric mean of speedups for 17 large graphs) over our implementation of the ParK k-core decomposition algorithm.

AB - k-core decomposition is a social network analytic that can be applied to centrality analysis, visualization, and community detection. There is a simple and well-known linear time sequential algorithm to perform k-core decomposition, and for this reason, k-core decomposition is widely used as a network analytic. In this work, we present a new shared-memory parallel algorithm called PKC for k-core decomposition on multicore platforms. This approach improves on the state-of-the-art implementations for k-core decomposition algorithms by reducing synchronization overhead and creating a smaller graph to process high-degree vertices. We show that PKC consistently outperforms implementations of other methods on a 32-core multicore server and on a collection of large sparse graphs. We achieve a 2.81× speedup (geometric mean of speedups for 17 large graphs) over our implementation of the ParK k-core decomposition algorithm.

UR - http://www.scopus.com/inward/record.url?scp=85028077884&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85028077884&partnerID=8YFLogxK

U2 - 10.1109/IPDPSW.2017.151

DO - 10.1109/IPDPSW.2017.151

M3 - Conference contribution

AN - SCOPUS:85028077884

T3 - Proceedings - 2017 IEEE 31st International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2017

SP - 1482

EP - 1491

BT - Proceedings - 2017 IEEE 31st International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2017

PB - Institute of Electrical and Electronics Engineers Inc.

ER -

Kabir H, Madduri K. Parallel k-Core decomposition on multicore platforms. In Proceedings - 2017 IEEE 31st International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2017. Institute of Electrical and Electronics Engineers Inc. 2017. p. 1482-1491. 7965211. (Proceedings - 2017 IEEE 31st International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2017). https://doi.org/10.1109/IPDPSW.2017.151