Far-Right Discourse on Migration

This dataset was collected using the TeleCatch application (Ruscica et al. 2024) and developed within the context of the “Uncovering visual narratives about human mobility on Telegram” project (Bia Carneiro et al. 2024). Data collection on the application works by pre-defining, based on previous research, a list of channels/groups where the collection will take place and drawing from a list of keywords of interest, in this case, related to migration, that will work as a target to retrieve the textual/ visual data shared in the selected groups. The final dataset consisted of 19.965 text messages retrieved from 24 groups/channels, in a universe of 1.085.474 messages shared in the Telegram channels since their creation. In addition, a total of 14,160 associated images were collected within these exchanges. It included information on message timestamps, channel id, username, text message, country, language, category, number of views each message received, media description, media type, author ID, name, username, reply to the author, and forwarded from the author. Captions for the images were generated using computer vision (Bootstrapped Language-Image Pretraining, (Li et al. 2022).