In Sweden, the growing influence of the radical right have turned cultural institutions into political symbols in an emerging ‘culture war’ (Harding, 2021). Carlsson, Hanell & Hansson (2022) show how digital forums and social media play a significant part in orchestrating the ideologically laden conflicts and confrontations that public libraries are currently facing. Digitally mediated threats from the radical right may obstruct the statutory mission of the institution to promote democracy, but knowledge about how such threats develop and unfold, as well as the relation between online interactions and offline events (Scrivens, Davies & Frank, 2020), is lacking. This prompts an investigation of the digital environments where supporters of radical right-wing ideas interact, share thoughts, construct collective identities, and coordinate actions towards libraries.

Methodologically, the pilot study explores user comments on online news websites (Landert 2014), applying a data-driven approach informed by sentiment analysis and topic modelling (TM) testing both conventional lexicon-based sentiment scoring and LDA (Latent Dirichlet Allocation) topic modelling, and more recent BERT (Bidirectional Encoder Representations from Transformers) modelling. For purposes of transparency and easy replicability, we carry out the analyses in the open-source environment KNIME (Konstanz Information Miner; see Ihrmark & Tyrkkö 2023).

The dataset of the pilot study contains 561 comments taken from 138 different articles published between 2019 and 2023 on the SamNytt website, the main media outlet for the Swedish radical right. The articles were selected by using the search term “bibliotek” in the site’s search function. The comments were made by 338 different user accounts, with a low number of individual accounts making more than 5 comments. An ongoing analysis suggests that the frequency and temporality of comments and upvotes provide a window into the processes shaping radical right-wing digital attacks that result in tangible consequences for public libraries.