Facebook and other social media are not to blame for filter bubbles, the problem of fragmentation is more fundamental

14/09/2020

Fragmentation in social networks which leads to the emergence of Echo Chambers have been a growing concern in recent years. Many have criticized Social Network sites for creating Filter Bubbles and therefore suggested that the solution must be in the hands of the engineers who design these platforms.

In their latest paper UCD School of Sociology's Taha Yasseri and the Oxford Internet Institute's Chris Blex show that the emergence of echo chambers is system-immanent and as long as individuals can freely and costlessly cut out their connections to people who have a different opinion, fragmentation will appear.

Moreover, they show that even if Social Network platforms, following the recommendations made by other scholars, utilize a positive algorithmic bias which would intentionally connect people of different opinions to one another, it will not solve the problem.

Instead, they propose that the introduction of a type of friction against cutting connections only due to disagreement, might help to prevent polarization and fragmentation in our social networks. In making this suggestion, they rely on examples of successful collaborative systems such as Wikipedia, where people of different opinions must keep working together and this mechanism not only prevents fragmentation, but also helps them to get closer to one another.

The paper, Positive algorithmic bias cannot stop fragmentation in homophilic networks, is published in The Journal of Mathematical Sociology.