Lagrangian relaxations for multiple network alignment

Eric Malmi, Sanjay Chawla, Aristides Gionis

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

We propose a principled approach for the problem of aligning multiple partially overlapping networks. The objective is to map multiple graphs into a single graph while preserving vertex and edge similarities. The problem is inspired by the task of integrating partial views of a family tree (genealogical network) into one unified network, but it also has applications, for example, in social and biological networks. Our approach, called Flan, introduces the idea of generalizing the facility location problem by adding a non-linear term to capture edge similarities and to infer the underlying entity network. The problem is solved using an alternating optimization procedure with a Lagrangian relaxation. Flan has the advantage of being able to leverage prior information on the number of entities, so that when this information is available, Flan is shown to work robustly without the need to use any ground truth data for fine-tuning method parameters. Additionally, we present three multiple-network extensions to an existing state-of-the-art pairwise alignment method called Natalie. Extensive experiments on synthetic, as well as real-world datasets on social networks and genealogical networks, attest to the effectiveness of the proposed approaches which clearly outperform a popular multiple network alignment method called IsoRankN.

Original languageEnglish
Pages (from-to)1331-1358
Number of pages28
JournalData Mining and Knowledge Discovery
Volume31
Issue number5
DOIs
Publication statusPublished - 1 Sep 2017

Fingerprint

Tuning
Experiments

Keywords

  • Facility location
  • Genealogical trees
  • Lagrangian relaxation
  • Multiple network alignment
  • Social networks

ASJC Scopus subject areas

  • Information Systems
  • Computer Science Applications
  • Computer Networks and Communications

Cite this

Lagrangian relaxations for multiple network alignment. / Malmi, Eric; Chawla, Sanjay; Gionis, Aristides.

In: Data Mining and Knowledge Discovery, Vol. 31, No. 5, 01.09.2017, p. 1331-1358.

Research output: Contribution to journalArticle

Malmi, Eric ; Chawla, Sanjay ; Gionis, Aristides. / Lagrangian relaxations for multiple network alignment. In: Data Mining and Knowledge Discovery. 2017 ; Vol. 31, No. 5. pp. 1331-1358.
@article{f0ce513a6a0948dc84867b1585bb61e8,
title = "Lagrangian relaxations for multiple network alignment",
abstract = "We propose a principled approach for the problem of aligning multiple partially overlapping networks. The objective is to map multiple graphs into a single graph while preserving vertex and edge similarities. The problem is inspired by the task of integrating partial views of a family tree (genealogical network) into one unified network, but it also has applications, for example, in social and biological networks. Our approach, called Flan, introduces the idea of generalizing the facility location problem by adding a non-linear term to capture edge similarities and to infer the underlying entity network. The problem is solved using an alternating optimization procedure with a Lagrangian relaxation. Flan has the advantage of being able to leverage prior information on the number of entities, so that when this information is available, Flan is shown to work robustly without the need to use any ground truth data for fine-tuning method parameters. Additionally, we present three multiple-network extensions to an existing state-of-the-art pairwise alignment method called Natalie. Extensive experiments on synthetic, as well as real-world datasets on social networks and genealogical networks, attest to the effectiveness of the proposed approaches which clearly outperform a popular multiple network alignment method called IsoRankN.",
keywords = "Facility location, Genealogical trees, Lagrangian relaxation, Multiple network alignment, Social networks",
author = "Eric Malmi and Sanjay Chawla and Aristides Gionis",
year = "2017",
month = "9",
day = "1",
doi = "10.1007/s10618-017-0505-2",
language = "English",
volume = "31",
pages = "1331--1358",
journal = "Data Mining and Knowledge Discovery",
issn = "1384-5810",
publisher = "Springer Netherlands",
number = "5",

}

TY - JOUR

T1 - Lagrangian relaxations for multiple network alignment

AU - Malmi, Eric

AU - Chawla, Sanjay

AU - Gionis, Aristides

PY - 2017/9/1

Y1 - 2017/9/1

N2 - We propose a principled approach for the problem of aligning multiple partially overlapping networks. The objective is to map multiple graphs into a single graph while preserving vertex and edge similarities. The problem is inspired by the task of integrating partial views of a family tree (genealogical network) into one unified network, but it also has applications, for example, in social and biological networks. Our approach, called Flan, introduces the idea of generalizing the facility location problem by adding a non-linear term to capture edge similarities and to infer the underlying entity network. The problem is solved using an alternating optimization procedure with a Lagrangian relaxation. Flan has the advantage of being able to leverage prior information on the number of entities, so that when this information is available, Flan is shown to work robustly without the need to use any ground truth data for fine-tuning method parameters. Additionally, we present three multiple-network extensions to an existing state-of-the-art pairwise alignment method called Natalie. Extensive experiments on synthetic, as well as real-world datasets on social networks and genealogical networks, attest to the effectiveness of the proposed approaches which clearly outperform a popular multiple network alignment method called IsoRankN.

AB - We propose a principled approach for the problem of aligning multiple partially overlapping networks. The objective is to map multiple graphs into a single graph while preserving vertex and edge similarities. The problem is inspired by the task of integrating partial views of a family tree (genealogical network) into one unified network, but it also has applications, for example, in social and biological networks. Our approach, called Flan, introduces the idea of generalizing the facility location problem by adding a non-linear term to capture edge similarities and to infer the underlying entity network. The problem is solved using an alternating optimization procedure with a Lagrangian relaxation. Flan has the advantage of being able to leverage prior information on the number of entities, so that when this information is available, Flan is shown to work robustly without the need to use any ground truth data for fine-tuning method parameters. Additionally, we present three multiple-network extensions to an existing state-of-the-art pairwise alignment method called Natalie. Extensive experiments on synthetic, as well as real-world datasets on social networks and genealogical networks, attest to the effectiveness of the proposed approaches which clearly outperform a popular multiple network alignment method called IsoRankN.

KW - Facility location

KW - Genealogical trees

KW - Lagrangian relaxation

KW - Multiple network alignment

KW - Social networks

UR - http://www.scopus.com/inward/record.url?scp=85016493281&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85016493281&partnerID=8YFLogxK

U2 - 10.1007/s10618-017-0505-2

DO - 10.1007/s10618-017-0505-2

M3 - Article

VL - 31

SP - 1331

EP - 1358

JO - Data Mining and Knowledge Discovery

JF - Data Mining and Knowledge Discovery

SN - 1384-5810

IS - 5

ER -