The complex link between filter bubbles and opinion polarization

Tracking #: 671-1651

Responsible editor: 

Victor de Boer

Submission Type: 

Position Paper


There is public and scholarly debate about the effects of personalized recommender systems implemented in online social networks, online markets, and search engines. On the one hand, it has been warned that personalization algorithms reduce the diversity of information diets which confirms users’ previously held attitudes and beliefs. Opinionated social media posts, shared news items, and online discussion could fragment social groups, alienate users with different political views, and ultimately foster opinion polarization. On the other hand, critics of this “personalization-polarization hypothesis” argue that the effects of personalization algorithms on information diets are too weak to have meaningful effects. Here, we argue that contributions to both sides of the debate fail to consider the complexity that arises when large numbers of interdependent Internet users interact and exert influence on one another in algorithmically governed communication systems. Reviewing insights from the literature of opinion dynamics in social networks, we demonstrate that opinion dynamics can be critically influenced by mechanisms active on three levels of analysis: the individual, local, and global level. We show that theoretical and empirical research on these three levels is needed to answer the question whether personalization fosters polarization or not, advocating an approach that combines rigorous theoretical modeling with the emergent field of data science.


Supplementary Files (optional): 

Previous Version: 


  • Reviewed

Data repository URLs:

Date of Submission: 

Monday, December 21, 2020

Date of Decision: 

Wednesday, April 7, 2021

Nanopublication URLs:



Solicited Reviews:

1 Comment

Meta-Review by Editor

Among the three reviewers, there is still considerable disagreement, specifically on the appropriateness of this article for this specific journal (in the form of a position paper). I have carefully considered the arguments by original reviewer 1 echoed and repeated by new reviewer 2 on one side, and the positive reviews on the other side. I would argue that this paper is indeed in scope of the journal, that the authors have done a commendable job in addressing the issues presented by the reviewers. I think the article is a valuable addition to the journal as a position paper and therefore decide to accept the paper. 

We do ask the authors to please take the reviews in this second round into account when preparing the final version. Specifically we ask 
- to use the suggestions (typos and terminology) by reviewer 1 to improve the paper
- to consider including a comment on the suggestions made by reviewer 3 
- to consider addressing the concern raised by reviewer 2: "The major missing ingredient is a conclusion that states: this is what data scientists can learn from these insights from taking a social science perspective." Even though to a certain extent this is written in the conclusion section, a suggestion is to more directly speak to the data scientist readers of this paper and list what can be learned. 

Victor de Boer (