Exploiting small world property for network clustering

Tieyun Qian, Qing Li, Jaideep Srivastava, Zhiyong Peng, Yang Yang, Shuo Wang

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

Graph partitioning is a traditional problem with many applications and a number of high-quality algorithms have been developed. Recently, demand for social network analysis arouses the new research interest on graph partitioning/clustering. Social networks differ from conventional graphs in that they exhibit some key properties like power-law and small-world property. Currently, these features are largely neglected in popular partitioning algorithms. In this paper, we present a novel framework which leverages the small-world property for finding clusters in social networks. The framework consists of several key features. Firstly, we define a total order, which combines the edge weight, the small-world weight, and the hub value, to better reflect the connection strength between two vertices. Secondly, we design a strategy using this ordered list, to greedily, yet effectively, refine existing partitioning algorithms for common objective functions. Thirdly, the proposed method is independent of the original approach, such that it could be integrated with any types of existing graph clustering algorithms. We conduct an extensive performance study on both real-life and synthetic datasets. The empirical results clearly demonstrate that our framework significantly improves the output of the state-of-the-art methods. Furthermore, we show that the proposed method returns clusters with both internal and external higher qualities.

Original languageEnglish
Pages (from-to)405-425
Number of pages21
JournalWorld Wide Web
Volume17
Issue number3
DOIs
Publication statusPublished - 1 May 2014
Externally publishedYes

Fingerprint

Electric network analysis
Clustering algorithms

Keywords

  • graph partitioning
  • network clustering
  • small world property

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Hardware and Architecture
  • Software

Cite this

Exploiting small world property for network clustering. / Qian, Tieyun; Li, Qing; Srivastava, Jaideep; Peng, Zhiyong; Yang, Yang; Wang, Shuo.

In: World Wide Web, Vol. 17, No. 3, 01.05.2014, p. 405-425.

Research output: Contribution to journalArticle

Qian, T, Li, Q, Srivastava, J, Peng, Z, Yang, Y & Wang, S 2014, 'Exploiting small world property for network clustering', World Wide Web, vol. 17, no. 3, pp. 405-425. https://doi.org/10.1007/s11280-013-0209-5
Qian, Tieyun ; Li, Qing ; Srivastava, Jaideep ; Peng, Zhiyong ; Yang, Yang ; Wang, Shuo. / Exploiting small world property for network clustering. In: World Wide Web. 2014 ; Vol. 17, No. 3. pp. 405-425.
@article{fddc10451743415d9cb45a1193d1d039,
title = "Exploiting small world property for network clustering",
abstract = "Graph partitioning is a traditional problem with many applications and a number of high-quality algorithms have been developed. Recently, demand for social network analysis arouses the new research interest on graph partitioning/clustering. Social networks differ from conventional graphs in that they exhibit some key properties like power-law and small-world property. Currently, these features are largely neglected in popular partitioning algorithms. In this paper, we present a novel framework which leverages the small-world property for finding clusters in social networks. The framework consists of several key features. Firstly, we define a total order, which combines the edge weight, the small-world weight, and the hub value, to better reflect the connection strength between two vertices. Secondly, we design a strategy using this ordered list, to greedily, yet effectively, refine existing partitioning algorithms for common objective functions. Thirdly, the proposed method is independent of the original approach, such that it could be integrated with any types of existing graph clustering algorithms. We conduct an extensive performance study on both real-life and synthetic datasets. The empirical results clearly demonstrate that our framework significantly improves the output of the state-of-the-art methods. Furthermore, we show that the proposed method returns clusters with both internal and external higher qualities.",
keywords = "graph partitioning, network clustering, small world property",
author = "Tieyun Qian and Qing Li and Jaideep Srivastava and Zhiyong Peng and Yang Yang and Shuo Wang",
year = "2014",
month = "5",
day = "1",
doi = "10.1007/s11280-013-0209-5",
language = "English",
volume = "17",
pages = "405--425",
journal = "World Wide Web",
issn = "1386-145X",
publisher = "Springer New York",
number = "3",

}

TY - JOUR

T1 - Exploiting small world property for network clustering

AU - Qian, Tieyun

AU - Li, Qing

AU - Srivastava, Jaideep

AU - Peng, Zhiyong

AU - Yang, Yang

AU - Wang, Shuo

PY - 2014/5/1

Y1 - 2014/5/1

N2 - Graph partitioning is a traditional problem with many applications and a number of high-quality algorithms have been developed. Recently, demand for social network analysis arouses the new research interest on graph partitioning/clustering. Social networks differ from conventional graphs in that they exhibit some key properties like power-law and small-world property. Currently, these features are largely neglected in popular partitioning algorithms. In this paper, we present a novel framework which leverages the small-world property for finding clusters in social networks. The framework consists of several key features. Firstly, we define a total order, which combines the edge weight, the small-world weight, and the hub value, to better reflect the connection strength between two vertices. Secondly, we design a strategy using this ordered list, to greedily, yet effectively, refine existing partitioning algorithms for common objective functions. Thirdly, the proposed method is independent of the original approach, such that it could be integrated with any types of existing graph clustering algorithms. We conduct an extensive performance study on both real-life and synthetic datasets. The empirical results clearly demonstrate that our framework significantly improves the output of the state-of-the-art methods. Furthermore, we show that the proposed method returns clusters with both internal and external higher qualities.

AB - Graph partitioning is a traditional problem with many applications and a number of high-quality algorithms have been developed. Recently, demand for social network analysis arouses the new research interest on graph partitioning/clustering. Social networks differ from conventional graphs in that they exhibit some key properties like power-law and small-world property. Currently, these features are largely neglected in popular partitioning algorithms. In this paper, we present a novel framework which leverages the small-world property for finding clusters in social networks. The framework consists of several key features. Firstly, we define a total order, which combines the edge weight, the small-world weight, and the hub value, to better reflect the connection strength between two vertices. Secondly, we design a strategy using this ordered list, to greedily, yet effectively, refine existing partitioning algorithms for common objective functions. Thirdly, the proposed method is independent of the original approach, such that it could be integrated with any types of existing graph clustering algorithms. We conduct an extensive performance study on both real-life and synthetic datasets. The empirical results clearly demonstrate that our framework significantly improves the output of the state-of-the-art methods. Furthermore, we show that the proposed method returns clusters with both internal and external higher qualities.

KW - graph partitioning

KW - network clustering

KW - small world property

UR - http://www.scopus.com/inward/record.url?scp=84919666068&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84919666068&partnerID=8YFLogxK

U2 - 10.1007/s11280-013-0209-5

DO - 10.1007/s11280-013-0209-5

M3 - Article

VL - 17

SP - 405

EP - 425

JO - World Wide Web

JF - World Wide Web

SN - 1386-145X

IS - 3

ER -