What is it about?
A challenge in social network analysis, is understanding the position (or stance) of people on a large set of topics. This work explores opportunities to infer topic and agreement correlations from online user interactions. Indeed, disagreement on one topic may make disagreement (or agreement) more likely for related topics. We propose a method able to find topic correlations in an unsupervised fashion, and predict the stance of an individual on topics for which their engagement has never been observed.
Featured Image
Photo by British Library on Unsplash
Why is it important?
The novelty comes from the simplicity yet effectiveness of the way our method combines social graph data to unveil hidden topic and agreement patterns. Moreover, we open-source two social graph datasets which create new research opportunities for network and social scientists.
Perspectives
Understand topical stances better opens doors for polarisation & echo chambers mitigation, moderation and content recommendations.
John Pougué-Biyong
University of Oxford
Read the Original
This page is a summary of: Learning Stance Embeddings from Signed Social Graphs, February 2023, ACM (Association for Computing Machinery),
DOI: 10.1145/3539597.3570401.
You can read the full text:
Resources
Contributors
The following have contributed to this page
