A Multi-agent Model for Polarization Under Confirmation Bias in Social Networks

Mário S. Alvim; Bernardo Amorim; Sophia Knight; Santiago Quintero; Frank Valencia

doi:10.1007/978-3-030-78089-0_2

A Multi-agent Model for Polarization Under Confirmation Bias in Social Networks

Mário S. Alvim, Bernardo Amorim, Sophia Knight, Santiago Quintero, Frank Valencia

Computer Science (Duluth)

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

4 Scopus citations

Abstract

We describe a model for polarization in multi-agent systems based on Esteban and Ray’s standard measure of polarization from economics. Agents evolve by updating their beliefs (opinions) based on an underlying influence graph, as in the standard DeGroot model for social learning, but under a confirmation bias; i.e., a discounting of opinions of agents with dissimilar views. We show that even under this bias polarization eventually vanishes (converges to zero) if the influence graph is strongly-connected. If the influence graph is a regular symmetric circulation, we determine the unique belief value to which all agents converge. Our more insightful result establishes that, under some natural assumptions, if polarization does not eventually vanish then either there is a disconnected subgroup of agents, or some agent influences others more than she is influenced. We also show that polarization does not necessarily vanish in weakly-connected graphs under confirmation bias. We illustrate our model with a series of case studies and simulations, and show how it relates to the classic DeGroot model for social learning.

Original language	English (US)
Title of host publication	Formal Techniques for Distributed Objects, Components, and Systems - 41st IFIP WG 6.1 International Conference, FORTE 2021, Held as Part of the 16th International Federated Conference on Distributed Computing Techniques, DisCoTec 2021, Proceedings
Editors	Kirstin Peters, Tim A. Willemse
Publisher	Springer Science and Business Media Deutschland GmbH
Pages	22-41
Number of pages	20
ISBN (Print)	9783030780883
DOIs	https://doi.org/10.1007/978-3-030-78089-0_2
State	Published - 2021
Event	41st IFIP WG 6.1 International Conference on Formal Techniques for Distributed Objects, Components, and Systems, FORTE 2021 held as part of 16th International Federated Conference on Distributed Computing Techniques, DisCoTec 2021 - Virtual, Online Duration: Jun 14 2021 → Jun 18 2021

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	12719 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	41st IFIP WG 6.1 International Conference on Formal Techniques for Distributed Objects, Components, and Systems, FORTE 2021 held as part of 16th International Federated Conference on Distributed Computing Techniques, DisCoTec 2021
City	Virtual, Online
Period	6/14/21 → 6/18/21

Bibliographical note

Funding Information:
Mário S. Alvim and Bernardo Amorim were partially supported by CNPq, CAPES and FAPEMIG. Santiago Quintero and Frank Valencia were partially supported by the ECOS-NORD project FACTS (C19M03).

Publisher Copyright:
© 2021, IFIP International Federation for Information Processing.

Keywords

Confirmation bias
Multi-agent systems
Polarization
Social networks

Access

10.1007/978-3-030-78089-0_2

OpenUrl availability

Full text

Cite this

Alvim, M. S., Amorim, B., Knight, S., Quintero, S., & Valencia, F. (2021). A Multi-agent Model for Polarization Under Confirmation Bias in Social Networks. In K. Peters, & T. A. Willemse (Eds.), Formal Techniques for Distributed Objects, Components, and Systems - 41st IFIP WG 6.1 International Conference, FORTE 2021, Held as Part of the 16th International Federated Conference on Distributed Computing Techniques, DisCoTec 2021, Proceedings (pp. 22-41). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 12719 LNCS). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-030-78089-0_2

A Multi-agent Model for Polarization Under Confirmation Bias in Social Networks. / Alvim, Mário S.; Amorim, Bernardo; Knight, Sophia et al.
Formal Techniques for Distributed Objects, Components, and Systems - 41st IFIP WG 6.1 International Conference, FORTE 2021, Held as Part of the 16th International Federated Conference on Distributed Computing Techniques, DisCoTec 2021, Proceedings. ed. / Kirstin Peters; Tim A. Willemse. Springer Science and Business Media Deutschland GmbH, 2021. p. 22-41 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 12719 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Alvim, MS, Amorim, B, Knight, S, Quintero, S & Valencia, F 2021, A Multi-agent Model for Polarization Under Confirmation Bias in Social Networks. in K Peters & TA Willemse (eds), Formal Techniques for Distributed Objects, Components, and Systems - 41st IFIP WG 6.1 International Conference, FORTE 2021, Held as Part of the 16th International Federated Conference on Distributed Computing Techniques, DisCoTec 2021, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 12719 LNCS, Springer Science and Business Media Deutschland GmbH, pp. 22-41, 41st IFIP WG 6.1 International Conference on Formal Techniques for Distributed Objects, Components, and Systems, FORTE 2021 held as part of 16th International Federated Conference on Distributed Computing Techniques, DisCoTec 2021, Virtual, Online, 6/14/21. https://doi.org/10.1007/978-3-030-78089-0_2

Alvim MS, Amorim B, Knight S, Quintero S, Valencia F. A Multi-agent Model for Polarization Under Confirmation Bias in Social Networks. In Peters K, Willemse TA, editors, Formal Techniques for Distributed Objects, Components, and Systems - 41st IFIP WG 6.1 International Conference, FORTE 2021, Held as Part of the 16th International Federated Conference on Distributed Computing Techniques, DisCoTec 2021, Proceedings. Springer Science and Business Media Deutschland GmbH. 2021. p. 22-41. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-030-78089-0_2

Alvim, Mário S. ; Amorim, Bernardo ; Knight, Sophia et al. / A Multi-agent Model for Polarization Under Confirmation Bias in Social Networks. Formal Techniques for Distributed Objects, Components, and Systems - 41st IFIP WG 6.1 International Conference, FORTE 2021, Held as Part of the 16th International Federated Conference on Distributed Computing Techniques, DisCoTec 2021, Proceedings. editor / Kirstin Peters ; Tim A. Willemse. Springer Science and Business Media Deutschland GmbH, 2021. pp. 22-41 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{35d0bd81044a40fababbe03f6f0978a2,

title = "A Multi-agent Model for Polarization Under Confirmation Bias in Social Networks",

abstract = "We describe a model for polarization in multi-agent systems based on Esteban and Ray{\textquoteright}s standard measure of polarization from economics. Agents evolve by updating their beliefs (opinions) based on an underlying influence graph, as in the standard DeGroot model for social learning, but under a confirmation bias; i.e., a discounting of opinions of agents with dissimilar views. We show that even under this bias polarization eventually vanishes (converges to zero) if the influence graph is strongly-connected. If the influence graph is a regular symmetric circulation, we determine the unique belief value to which all agents converge. Our more insightful result establishes that, under some natural assumptions, if polarization does not eventually vanish then either there is a disconnected subgroup of agents, or some agent influences others more than she is influenced. We also show that polarization does not necessarily vanish in weakly-connected graphs under confirmation bias. We illustrate our model with a series of case studies and simulations, and show how it relates to the classic DeGroot model for social learning.",

keywords = "Confirmation bias, Multi-agent systems, Polarization, Social networks",

author = "Alvim, {M{\'a}rio S.} and Bernardo Amorim and Sophia Knight and Santiago Quintero and Frank Valencia",

note = "Funding Information: M{\'a}rio S. Alvim and Bernardo Amorim were partially supported by CNPq, CAPES and FAPEMIG. Santiago Quintero and Frank Valencia were partially supported by the ECOS-NORD project FACTS (C19M03). Publisher Copyright: {\textcopyright} 2021, IFIP International Federation for Information Processing.; 41st IFIP WG 6.1 International Conference on Formal Techniques for Distributed Objects, Components, and Systems, FORTE 2021 held as part of 16th International Federated Conference on Distributed Computing Techniques, DisCoTec 2021 ; Conference date: 14-06-2021 Through 18-06-2021",

year = "2021",

doi = "10.1007/978-3-030-78089-0_2",

language = "English (US)",

isbn = "9783030780883",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Science and Business Media Deutschland GmbH",

pages = "22--41",

editor = "Kirstin Peters and Willemse, {Tim A.}",

booktitle = "Formal Techniques for Distributed Objects, Components, and Systems - 41st IFIP WG 6.1 International Conference, FORTE 2021, Held as Part of the 16th International Federated Conference on Distributed Computing Techniques, DisCoTec 2021, Proceedings",

address = "Germany",

}

TY - GEN

T1 - A Multi-agent Model for Polarization Under Confirmation Bias in Social Networks

AU - Alvim, Mário S.

AU - Amorim, Bernardo

AU - Knight, Sophia

AU - Quintero, Santiago

AU - Valencia, Frank

N1 - Funding Information: Mário S. Alvim and Bernardo Amorim were partially supported by CNPq, CAPES and FAPEMIG. Santiago Quintero and Frank Valencia were partially supported by the ECOS-NORD project FACTS (C19M03). Publisher Copyright: © 2021, IFIP International Federation for Information Processing.

PY - 2021

Y1 - 2021

N2 - We describe a model for polarization in multi-agent systems based on Esteban and Ray’s standard measure of polarization from economics. Agents evolve by updating their beliefs (opinions) based on an underlying influence graph, as in the standard DeGroot model for social learning, but under a confirmation bias; i.e., a discounting of opinions of agents with dissimilar views. We show that even under this bias polarization eventually vanishes (converges to zero) if the influence graph is strongly-connected. If the influence graph is a regular symmetric circulation, we determine the unique belief value to which all agents converge. Our more insightful result establishes that, under some natural assumptions, if polarization does not eventually vanish then either there is a disconnected subgroup of agents, or some agent influences others more than she is influenced. We also show that polarization does not necessarily vanish in weakly-connected graphs under confirmation bias. We illustrate our model with a series of case studies and simulations, and show how it relates to the classic DeGroot model for social learning.

AB - We describe a model for polarization in multi-agent systems based on Esteban and Ray’s standard measure of polarization from economics. Agents evolve by updating their beliefs (opinions) based on an underlying influence graph, as in the standard DeGroot model for social learning, but under a confirmation bias; i.e., a discounting of opinions of agents with dissimilar views. We show that even under this bias polarization eventually vanishes (converges to zero) if the influence graph is strongly-connected. If the influence graph is a regular symmetric circulation, we determine the unique belief value to which all agents converge. Our more insightful result establishes that, under some natural assumptions, if polarization does not eventually vanish then either there is a disconnected subgroup of agents, or some agent influences others more than she is influenced. We also show that polarization does not necessarily vanish in weakly-connected graphs under confirmation bias. We illustrate our model with a series of case studies and simulations, and show how it relates to the classic DeGroot model for social learning.

KW - Confirmation bias

KW - Multi-agent systems

KW - Polarization

KW - Social networks

UR - http://www.scopus.com/inward/record.url?scp=85111373559&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85111373559&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-78089-0_2

DO - 10.1007/978-3-030-78089-0_2

M3 - Conference contribution

AN - SCOPUS:85111373559

SN - 9783030780883

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 22

EP - 41

BT - Formal Techniques for Distributed Objects, Components, and Systems - 41st IFIP WG 6.1 International Conference, FORTE 2021, Held as Part of the 16th International Federated Conference on Distributed Computing Techniques, DisCoTec 2021, Proceedings

A2 - Peters, Kirstin

A2 - Willemse, Tim A.

PB - Springer Science and Business Media Deutschland GmbH

T2 - 41st IFIP WG 6.1 International Conference on Formal Techniques for Distributed Objects, Components, and Systems, FORTE 2021 held as part of 16th International Federated Conference on Distributed Computing Techniques, DisCoTec 2021

Y2 - 14 June 2021 through 18 June 2021

ER -

A Multi-agent Model for Polarization Under Confirmation Bias in Social Networks

Abstract

Publication series

Conference

Bibliographical note

Keywords

Access

OpenUrl availability

Other files and links

Fingerprint

Cite this