Abstract
In this paper we develop a preliminary model for social networks, and a measure of the level of polarization in these social networks, based on Esteban and Ray’s classic measure of polarization for economic situations. Our model includes information about each agent’s quantitative strength of belief in a proposition of interest and a representation of the strength of each agent’s influence on every other agent. We consider how the model changes over time as agents interact and communicate, and include several different options for belief update, including rational belief update and update taking into account irrational responses such as confirmation bias and the backfire effect. Under various scenarios, we consider the evolution of polarization over time, and the implications of these results for real world social networks.
Original language | English (US) |
---|---|
Title of host publication | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
Publisher | Springer Verlag |
Pages | 419-441 |
Number of pages | 23 |
DOIs | |
State | Published - 2019 |
Publication series
Name | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
---|---|
Volume | 11760 LNCS |
ISSN (Print) | 0302-9743 |
ISSN (Electronic) | 1611-3349 |
Bibliographical note
Funding Information:Mário S. Alvim was partially supported by the National Council for Scientific and Technological Development (CNPq), CAPES, and FAPEMIG. Frank Valencia has been partially supported by the ECOS-NORD project FACTS (C19M03).
Publisher Copyright:
© 2019, Springer Nature Switzerland AG.