A guide to selecting a network similarity method

Sucheta Soundarajan, Tina Eliassi-Rad, Brian Gallagher

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

16 Citations (Scopus)

Abstract

We consider the problem of determining how similar two networks (without known node-correspondences) are. This problem occurs frequently in real-world applications such as transfer learning and change detection. Many networksimilarity methods exist; and it is unclear how one should select from amongst them. We provide the first empirical study on the relationships between different networksimilarity methods. Specifically, we present (1) an approach for identifying groups of comparable network-similarity methods and (2) an approach for computing the consensus among a given set of network-similarity methods. We compare and contrast twenty network-similarity methods by applying our approaches to a variety of real datasets spanning multiple domains. Our experiments demonstrate that (1) different network-similarity methods are surprisingly well correlated, (2) some complex network-similarity methods can be closely approximated by a much simpler method, and (3) a few network-similarity methods produce rankings that are very close to the consensus ranking.

Original languageEnglish (US)
Title of host publicationSIAM International Conference on Data Mining 2014, SDM 2014
PublisherSociety for Industrial and Applied Mathematics Publications
Pages1037-1045
Number of pages9
Volume2
ISBN (Print)9781510811515
DOIs
StatePublished - 2014
Externally publishedYes
Event14th SIAM International Conference on Data Mining, SDM 2014 - Philadelphia, United States
Duration: Apr 24 2014Apr 26 2014

Other

Other14th SIAM International Conference on Data Mining, SDM 2014
CountryUnited States
CityPhiladelphia
Period4/24/144/26/14

Fingerprint

Complex networks
Experiments

ASJC Scopus subject areas

  • Computer Science Applications
  • Software

Cite this

Soundarajan, S., Eliassi-Rad, T., & Gallagher, B. (2014). A guide to selecting a network similarity method. In SIAM International Conference on Data Mining 2014, SDM 2014 (Vol. 2, pp. 1037-1045). Society for Industrial and Applied Mathematics Publications. https://doi.org/10.1137/1.9781611973440.118

A guide to selecting a network similarity method. / Soundarajan, Sucheta; Eliassi-Rad, Tina; Gallagher, Brian.

SIAM International Conference on Data Mining 2014, SDM 2014. Vol. 2 Society for Industrial and Applied Mathematics Publications, 2014. p. 1037-1045.

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

Soundarajan, S, Eliassi-Rad, T & Gallagher, B 2014, A guide to selecting a network similarity method. in SIAM International Conference on Data Mining 2014, SDM 2014. vol. 2, Society for Industrial and Applied Mathematics Publications, pp. 1037-1045, 14th SIAM International Conference on Data Mining, SDM 2014, Philadelphia, United States, 4/24/14. https://doi.org/10.1137/1.9781611973440.118
Soundarajan S, Eliassi-Rad T, Gallagher B. A guide to selecting a network similarity method. In SIAM International Conference on Data Mining 2014, SDM 2014. Vol. 2. Society for Industrial and Applied Mathematics Publications. 2014. p. 1037-1045 https://doi.org/10.1137/1.9781611973440.118
Soundarajan, Sucheta ; Eliassi-Rad, Tina ; Gallagher, Brian. / A guide to selecting a network similarity method. SIAM International Conference on Data Mining 2014, SDM 2014. Vol. 2 Society for Industrial and Applied Mathematics Publications, 2014. pp. 1037-1045
@inproceedings{1a1172c66e954ac4ad5fb20a644099cd,
title = "A guide to selecting a network similarity method",
abstract = "We consider the problem of determining how similar two networks (without known node-correspondences) are. This problem occurs frequently in real-world applications such as transfer learning and change detection. Many networksimilarity methods exist; and it is unclear how one should select from amongst them. We provide the first empirical study on the relationships between different networksimilarity methods. Specifically, we present (1) an approach for identifying groups of comparable network-similarity methods and (2) an approach for computing the consensus among a given set of network-similarity methods. We compare and contrast twenty network-similarity methods by applying our approaches to a variety of real datasets spanning multiple domains. Our experiments demonstrate that (1) different network-similarity methods are surprisingly well correlated, (2) some complex network-similarity methods can be closely approximated by a much simpler method, and (3) a few network-similarity methods produce rankings that are very close to the consensus ranking.",
author = "Sucheta Soundarajan and Tina Eliassi-Rad and Brian Gallagher",
year = "2014",
doi = "10.1137/1.9781611973440.118",
language = "English (US)",
isbn = "9781510811515",
volume = "2",
pages = "1037--1045",
booktitle = "SIAM International Conference on Data Mining 2014, SDM 2014",
publisher = "Society for Industrial and Applied Mathematics Publications",

}

TY - GEN

T1 - A guide to selecting a network similarity method

AU - Soundarajan, Sucheta

AU - Eliassi-Rad, Tina

AU - Gallagher, Brian

PY - 2014

Y1 - 2014

N2 - We consider the problem of determining how similar two networks (without known node-correspondences) are. This problem occurs frequently in real-world applications such as transfer learning and change detection. Many networksimilarity methods exist; and it is unclear how one should select from amongst them. We provide the first empirical study on the relationships between different networksimilarity methods. Specifically, we present (1) an approach for identifying groups of comparable network-similarity methods and (2) an approach for computing the consensus among a given set of network-similarity methods. We compare and contrast twenty network-similarity methods by applying our approaches to a variety of real datasets spanning multiple domains. Our experiments demonstrate that (1) different network-similarity methods are surprisingly well correlated, (2) some complex network-similarity methods can be closely approximated by a much simpler method, and (3) a few network-similarity methods produce rankings that are very close to the consensus ranking.

AB - We consider the problem of determining how similar two networks (without known node-correspondences) are. This problem occurs frequently in real-world applications such as transfer learning and change detection. Many networksimilarity methods exist; and it is unclear how one should select from amongst them. We provide the first empirical study on the relationships between different networksimilarity methods. Specifically, we present (1) an approach for identifying groups of comparable network-similarity methods and (2) an approach for computing the consensus among a given set of network-similarity methods. We compare and contrast twenty network-similarity methods by applying our approaches to a variety of real datasets spanning multiple domains. Our experiments demonstrate that (1) different network-similarity methods are surprisingly well correlated, (2) some complex network-similarity methods can be closely approximated by a much simpler method, and (3) a few network-similarity methods produce rankings that are very close to the consensus ranking.

UR - http://www.scopus.com/inward/record.url?scp=84959911503&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84959911503&partnerID=8YFLogxK

U2 - 10.1137/1.9781611973440.118

DO - 10.1137/1.9781611973440.118

M3 - Conference contribution

AN - SCOPUS:84959911503

SN - 9781510811515

VL - 2

SP - 1037

EP - 1045

BT - SIAM International Conference on Data Mining 2014, SDM 2014

PB - Society for Industrial and Applied Mathematics Publications

ER -