Berlin 2024 – wissenschaftliches Programm
Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
SOE: Fachverband Physik sozio-ökonomischer Systeme
SOE 3: Machine Learning
SOE 3.1: Vortrag
Montag, 18. März 2024, 12:15–12:30, MA 001
Mapping news sharing on Twitter: A bottom-up approach based on network embeddings — •Felix Gaisbauer1, Armin Pournaki2,3, and Jakob Ohme1 — 1Weizenbaum-Institut für die vernetzte Gesellschaft e.V. — 2Max-Planck-Institut für Mathematik in den Naturwissenschaften — 3medialab, Sciences Po
News sharing on digital platforms is a crucial activity that determines the digital spaces millions of users navigate. Yet, we know little about general patterns of news sharing. We utilize a combination of three data sources - which we combine via network embedding methods and automated text analysis - to elucidate the extent to which sharing patterns of certain political user groups consist of specific outlets/topics/articles or have unknown diversity.
We collected all tweets which contained a link to one of 26 legacy or alternative news outlets for March 2023 (2.5M tweets). The full texts of the articles were crawled if available (30K texts); articles were assigned topics with a paragraph-based BERTopic model. The follower network of German MPs was also collected. This was used to embed followers and MPs in a latent political space using correspondence analysis.
This allows to investigate which types of articles are shared in which political region(s) of the latent space. To explore this systematically, we apply measures of collective sharing breadth and depth in the embeddings with respect to specific outlets, topics or single news events. All in all, this enables a previously unexplored bottom-up view on news sharing on Twitter.
Keywords: Networks; Twitter; Embeddings