Linguistic factors that influence the diffusion of information in online social networks

Current project leader: Lejla Džanko

Description: Whether we want to help propagate useful ideas and grow online collective moments to see a positive change in the world or to stop false narratives from catching on, it is crucial to study the mechanisms by which information spreads – the diffusion of information – in online social networks. In this research project, we set out to find if there is a way a message could be framed that would help it go viral above and beyond other factors, such as the popularity of the message source (here, a social media user) or the network configuration. Our goal is to test if the linguistic features previously identified in the literature are positive predictors of message diffusion in the context of social movements on Twitter. Additionally, we are interested in positive predictors of the messages’ out-group (vs in-group) popularity. To address these goals, we will conduct automated text analysis on a large corpus of tweets on different social movements. We will also apply network analysis, to identify the structure and parameters of the user network. To confirm our findings, we will test linguistic factors which show the strongest positive influence on information diffusion in an experimental study. 


  • Polish National Science Foundation Grant 2017/27/B/HS6/01049. Social Grammar Model – basic and applied mechanisms – awarded to Magdalena Formanowicz (2018-2021).