Stopping fake news: Who should be banned?

• Published in: International journal of data science and analytics 2024-09, Vol.18 (3), p.259-273
• Main Authors: Fierens, Pablo Ignacio, Rêgo, Leandro Chaves
• Format: Article
• Language: English
• Subjects: Artificial Intelligence; Business Information Systems; Computational Biology/Bioinformatics; Computer Science; Data Mining and Knowledge Discovery; Database Management; Regular Paper
• ISSN: 2364-415X; 2364-4168
• DOI: 10.1007/s41060-024-00532-x

Fake news and misinformation spread in online social networks in a manner similar to contagious diseases. One possibility to thwart the contagion cascade is to selectively remove a small number of nodes from the network. Although most of the literature has focused on the selection of those nodes on the basis of their topological position in the network, we pose that attributes of the nodes themselves can be more relevant in certain situations. In order to demonstrate this hypothesis, we introduce a new model of news propagation that accounts for nodes’ attributes. In particular, we introduce three important characteristics of a node: the influence capacity, the resistance to be influenced and the resistance to become an information spreader. Besides offering an intuitive justification for the model and these new parameters, we relate them to other proposals in the literature. Under the new model and using numerical simulations on both synthetic and real life networks, we show that nodes’ attributes can be more important than their graph structural properties in choosing an adequate set of vertices to be removed with the purpose of mitigating fake news propagation. Furthermore, our results suggest that removal of nodes with high influence power is more effective in denser networks and when the influence of a few nodes is much larger than that of the general population.