GRAFT: An approximate graphlet counting algorithm for large graph analysis

Mahmudur Rahman, Mansurul Bhuiyan, Mohammad Al Hasan

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

12 Citations (Scopus)

Abstract

Graphlet frequency distribution (GFD) is an analysis tool for understanding the variance of local structure in a graph. Many recent works use GFD for comparing, and characterizing real-life networks. However, the main bottleneck for graph analysis using GFD is the excessive computation cost for obtaining the frequency of each of the graphlets in a large network. To overcome this, we propose a simple, yet powerful algorithm, called GRAFT, that obtains the approximate graphlet frequency for all graphlets that have upto 5 vertices. Comparing to an exact counting algorithm, our algorithm achieves a speedup factor between 10 and 100 for a negligible counting error, which is, on average, less than 5%; For example, exact graphlet counting for ca-AstroPh takes approximately 3 days; but, GRAFT runs for 45 minutes to perform the same task with a counting accuracy of 95.6%.

Original languageEnglish
Title of host publicationACM International Conference Proceeding Series
Pages1467-1471
Number of pages5
DOIs
Publication statusPublished - 19 Dec 2012
Event21st ACM International Conference on Information and Knowledge Management, CIKM 2012 - Maui, HI, United States
Duration: 29 Oct 20122 Nov 2012

Other

Other21st ACM International Conference on Information and Knowledge Management, CIKM 2012
CountryUnited States
CityMaui, HI
Period29/10/122/11/12

Fingerprint

Costs

Keywords

  • approximate graphlet counting
  • graph analysis
  • graphlet frequency distribution

ASJC Scopus subject areas

  • Human-Computer Interaction
  • Computer Networks and Communications
  • Computer Vision and Pattern Recognition
  • Software

Cite this

Rahman, M., Bhuiyan, M., & Hasan, M. A. (2012). GRAFT: An approximate graphlet counting algorithm for large graph analysis. In ACM International Conference Proceeding Series (pp. 1467-1471) https://doi.org/10.1145/2396761.2398454

GRAFT : An approximate graphlet counting algorithm for large graph analysis. / Rahman, Mahmudur; Bhuiyan, Mansurul; Hasan, Mohammad Al.

ACM International Conference Proceeding Series. 2012. p. 1467-1471.

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

Rahman, M, Bhuiyan, M & Hasan, MA 2012, GRAFT: An approximate graphlet counting algorithm for large graph analysis. in ACM International Conference Proceeding Series. pp. 1467-1471, 21st ACM International Conference on Information and Knowledge Management, CIKM 2012, Maui, HI, United States, 29/10/12. https://doi.org/10.1145/2396761.2398454
Rahman M, Bhuiyan M, Hasan MA. GRAFT: An approximate graphlet counting algorithm for large graph analysis. In ACM International Conference Proceeding Series. 2012. p. 1467-1471 https://doi.org/10.1145/2396761.2398454
Rahman, Mahmudur ; Bhuiyan, Mansurul ; Hasan, Mohammad Al. / GRAFT : An approximate graphlet counting algorithm for large graph analysis. ACM International Conference Proceeding Series. 2012. pp. 1467-1471
@inproceedings{027f5b3f1d674aa3861d8d6a76e01be2,
title = "GRAFT: An approximate graphlet counting algorithm for large graph analysis",
abstract = "Graphlet frequency distribution (GFD) is an analysis tool for understanding the variance of local structure in a graph. Many recent works use GFD for comparing, and characterizing real-life networks. However, the main bottleneck for graph analysis using GFD is the excessive computation cost for obtaining the frequency of each of the graphlets in a large network. To overcome this, we propose a simple, yet powerful algorithm, called GRAFT, that obtains the approximate graphlet frequency for all graphlets that have upto 5 vertices. Comparing to an exact counting algorithm, our algorithm achieves a speedup factor between 10 and 100 for a negligible counting error, which is, on average, less than 5{\%}; For example, exact graphlet counting for ca-AstroPh takes approximately 3 days; but, GRAFT runs for 45 minutes to perform the same task with a counting accuracy of 95.6{\%}.",
keywords = "approximate graphlet counting, graph analysis, graphlet frequency distribution",
author = "Mahmudur Rahman and Mansurul Bhuiyan and Hasan, {Mohammad Al}",
year = "2012",
month = "12",
day = "19",
doi = "10.1145/2396761.2398454",
language = "English",
isbn = "9781450311564",
pages = "1467--1471",
booktitle = "ACM International Conference Proceeding Series",

}

TY - GEN

T1 - GRAFT

T2 - An approximate graphlet counting algorithm for large graph analysis

AU - Rahman, Mahmudur

AU - Bhuiyan, Mansurul

AU - Hasan, Mohammad Al

PY - 2012/12/19

Y1 - 2012/12/19

N2 - Graphlet frequency distribution (GFD) is an analysis tool for understanding the variance of local structure in a graph. Many recent works use GFD for comparing, and characterizing real-life networks. However, the main bottleneck for graph analysis using GFD is the excessive computation cost for obtaining the frequency of each of the graphlets in a large network. To overcome this, we propose a simple, yet powerful algorithm, called GRAFT, that obtains the approximate graphlet frequency for all graphlets that have upto 5 vertices. Comparing to an exact counting algorithm, our algorithm achieves a speedup factor between 10 and 100 for a negligible counting error, which is, on average, less than 5%; For example, exact graphlet counting for ca-AstroPh takes approximately 3 days; but, GRAFT runs for 45 minutes to perform the same task with a counting accuracy of 95.6%.

AB - Graphlet frequency distribution (GFD) is an analysis tool for understanding the variance of local structure in a graph. Many recent works use GFD for comparing, and characterizing real-life networks. However, the main bottleneck for graph analysis using GFD is the excessive computation cost for obtaining the frequency of each of the graphlets in a large network. To overcome this, we propose a simple, yet powerful algorithm, called GRAFT, that obtains the approximate graphlet frequency for all graphlets that have upto 5 vertices. Comparing to an exact counting algorithm, our algorithm achieves a speedup factor between 10 and 100 for a negligible counting error, which is, on average, less than 5%; For example, exact graphlet counting for ca-AstroPh takes approximately 3 days; but, GRAFT runs for 45 minutes to perform the same task with a counting accuracy of 95.6%.

KW - approximate graphlet counting

KW - graph analysis

KW - graphlet frequency distribution

UR - http://www.scopus.com/inward/record.url?scp=84871089313&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84871089313&partnerID=8YFLogxK

U2 - 10.1145/2396761.2398454

DO - 10.1145/2396761.2398454

M3 - Conference contribution

AN - SCOPUS:84871089313

SN - 9781450311564

SP - 1467

EP - 1471

BT - ACM International Conference Proceeding Series

ER -