Scientists at Aalto University in Finland propose an algorithmic solution to minimize societal polarization by linking people with opposing viewpoints and assessing them on Twitter.
Modeling user interactions around a given controversial subject on Twitter on an endorsement graph, with nodes representing Twitter users, tends to yield a strongly biclustered structure.
Aalto professor Aristides Gionis says the algorithm "can be applied on a large scale and is language- and domain-independent. The main algorithm is based on the finding that for a special type of network simulating a polarized network, the best bridges we can add to the network are between the nodes with the highest degrees on either side."
High-degree users are typically well-known and have numerous followers, and the algorithm's application to discussions on U.S. election results suggests "creating a bridge between @hillaryclinton and @breitbartnews would reduce polarization the most," says Aalto's Kiran Garimella.
From Aalto University
View Full Article
Abstracts Copyright © 2017 Information Inc., Bethesda, Maryland, USA
No entries found