Extreme-Right in Spain: An analysis of the internal communication of the VOX party through DeepLearning

This research focuses on the offer side of the Spanish far-right political party VOX, namely its internal organization and forms of communication between party members and sympathizers. This research analyzes the internal communication of the party VOX through its Telegram channels. VOX is the Spanish party with the most subscribers on Telegram, doubling the number of subscribers of the second most subscribed party, Podemos. 53 official channels of the party dedicated to different areas in Spain were analyzed using machine learning methodology due to the amount of data and the precision of the model. The information was collected through Telegram's API and Python libraries. Messages were analyzed using an open pre-trained model, BERTopic. The model works with more than 50 languages and the main functionality is to extract the main topics in a text corpus using the embedding technique and the c- TF-IDF algorithm to create dense topic clusters. The aim of the research is to relate the internal communication of the party to the internal democracy of the party. This research is particularly relevant because it aims to analyze not only what they say, but also how they strategically use the Telegram platform for their internal communications. The architecture of their Telegram channels and their main topics were analyzed from the perspective of the affordance of their channels. One of the main findings of this research is the use of the Telegram platform as an intermediate platform for the distribution of their messages on Twitter. The architecture of their Telegram channels reproduces their national ideology, regional, centralized, and with the absence of autonomous communities. This research shows that it is possible to identify important features of intra-party organization and intra-party democracy through the analysis of intra-party communication strategies.