Bayesian approach to multilayer stochastic blockmodel and network changepoint detection

Yunkyu Sohn, Jong Hee Park

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

Network scholars commonly encounter multiple networks, each of which is possibly governed by distinct generation rules while sharing a node group structure. Although the stochastic blockmodeling-detecting such latent group structures with group-specific connection profiles-has been a major topic of recent research, the focus has been given to the assortative group discovery of a single network. Despite its universality, concepts, and techniques for simultaneous characterization of node traits of multilayer networks, constructed by stacking multiple networks into layers, have been limited. Here, we propose a Bayesian multilayer stochastic blockmodeling framework that uncovers layer-common node traits and factors associated with layer-specific network generating functions. Without assuming a priori layer-specific generation rules, our fully Bayesian treatment allows probabilistic inference of latent traits. We extend the approach to detect changes in block structures embedded in temporal layers of network time series. We demonstrate the method using synthetic multilayer networks with assortative, disassortative, core-periphery, and overlapping community structures. Finally, we apply the method to empirical social network datasets, and find that it detects significant latent traits and structural changepoints. In particular, we uncover endogenous historical regimes associated with distinct constellations of states in United States Senate roll call vote similarity patterns.

Original languageEnglish
Pages (from-to)164-186
Number of pages23
JournalNetwork Science
Volume5
Issue number2
DOIs
Publication statusPublished - 2017 Jun 1
Externally publishedYes

Fingerprint

Bayes Theorem
Multilayers
Social Support
Network layers
Time series
Research
Group
senate
time series
voter
social network
regime
Datasets

Keywords

  • hidden Markov model
  • multilayer network
  • network changepoint detection
  • network time series
  • stochastic blockmodel
  • tensor decomposition

ASJC Scopus subject areas

  • Social Psychology
  • Communication
  • Sociology and Political Science

Cite this

Bayesian approach to multilayer stochastic blockmodel and network changepoint detection. / Sohn, Yunkyu; Park, Jong Hee.

In: Network Science, Vol. 5, No. 2, 01.06.2017, p. 164-186.

Research output: Contribution to journalArticle

@article{dea01ae06c1442fd8f8b95cd7e3ec056,
title = "Bayesian approach to multilayer stochastic blockmodel and network changepoint detection",
abstract = "Network scholars commonly encounter multiple networks, each of which is possibly governed by distinct generation rules while sharing a node group structure. Although the stochastic blockmodeling-detecting such latent group structures with group-specific connection profiles-has been a major topic of recent research, the focus has been given to the assortative group discovery of a single network. Despite its universality, concepts, and techniques for simultaneous characterization of node traits of multilayer networks, constructed by stacking multiple networks into layers, have been limited. Here, we propose a Bayesian multilayer stochastic blockmodeling framework that uncovers layer-common node traits and factors associated with layer-specific network generating functions. Without assuming a priori layer-specific generation rules, our fully Bayesian treatment allows probabilistic inference of latent traits. We extend the approach to detect changes in block structures embedded in temporal layers of network time series. We demonstrate the method using synthetic multilayer networks with assortative, disassortative, core-periphery, and overlapping community structures. Finally, we apply the method to empirical social network datasets, and find that it detects significant latent traits and structural changepoints. In particular, we uncover endogenous historical regimes associated with distinct constellations of states in United States Senate roll call vote similarity patterns.",
keywords = "hidden Markov model, multilayer network, network changepoint detection, network time series, stochastic blockmodel, tensor decomposition",
author = "Yunkyu Sohn and Park, {Jong Hee}",
year = "2017",
month = "6",
day = "1",
doi = "10.1017/nws.2017.13",
language = "English",
volume = "5",
pages = "164--186",
journal = "Network Science",
issn = "2050-1242",
publisher = "Cambridge University Press",
number = "2",

}

TY - JOUR

T1 - Bayesian approach to multilayer stochastic blockmodel and network changepoint detection

AU - Sohn, Yunkyu

AU - Park, Jong Hee

PY - 2017/6/1

Y1 - 2017/6/1

N2 - Network scholars commonly encounter multiple networks, each of which is possibly governed by distinct generation rules while sharing a node group structure. Although the stochastic blockmodeling-detecting such latent group structures with group-specific connection profiles-has been a major topic of recent research, the focus has been given to the assortative group discovery of a single network. Despite its universality, concepts, and techniques for simultaneous characterization of node traits of multilayer networks, constructed by stacking multiple networks into layers, have been limited. Here, we propose a Bayesian multilayer stochastic blockmodeling framework that uncovers layer-common node traits and factors associated with layer-specific network generating functions. Without assuming a priori layer-specific generation rules, our fully Bayesian treatment allows probabilistic inference of latent traits. We extend the approach to detect changes in block structures embedded in temporal layers of network time series. We demonstrate the method using synthetic multilayer networks with assortative, disassortative, core-periphery, and overlapping community structures. Finally, we apply the method to empirical social network datasets, and find that it detects significant latent traits and structural changepoints. In particular, we uncover endogenous historical regimes associated with distinct constellations of states in United States Senate roll call vote similarity patterns.

AB - Network scholars commonly encounter multiple networks, each of which is possibly governed by distinct generation rules while sharing a node group structure. Although the stochastic blockmodeling-detecting such latent group structures with group-specific connection profiles-has been a major topic of recent research, the focus has been given to the assortative group discovery of a single network. Despite its universality, concepts, and techniques for simultaneous characterization of node traits of multilayer networks, constructed by stacking multiple networks into layers, have been limited. Here, we propose a Bayesian multilayer stochastic blockmodeling framework that uncovers layer-common node traits and factors associated with layer-specific network generating functions. Without assuming a priori layer-specific generation rules, our fully Bayesian treatment allows probabilistic inference of latent traits. We extend the approach to detect changes in block structures embedded in temporal layers of network time series. We demonstrate the method using synthetic multilayer networks with assortative, disassortative, core-periphery, and overlapping community structures. Finally, we apply the method to empirical social network datasets, and find that it detects significant latent traits and structural changepoints. In particular, we uncover endogenous historical regimes associated with distinct constellations of states in United States Senate roll call vote similarity patterns.

KW - hidden Markov model

KW - multilayer network

KW - network changepoint detection

KW - network time series

KW - stochastic blockmodel

KW - tensor decomposition

UR - http://www.scopus.com/inward/record.url?scp=85020408520&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85020408520&partnerID=8YFLogxK

U2 - 10.1017/nws.2017.13

DO - 10.1017/nws.2017.13

M3 - Article

AN - SCOPUS:85020408520

VL - 5

SP - 164

EP - 186

JO - Network Science

JF - Network Science

SN - 2050-1242

IS - 2

ER -