The accelerated expansion of digital technology has undoubtedly impacted the dynamics of information production, distribution, and interpretation. Many news aggregator Telegram channels exist now solely to spread highly narratively distorted pieces of events, framed in a way to serve certain political, economic, and social interests. The current technological landscape, thus, pushes the need for computation processes that can allow researchers to carry out more efficient textual analyses on larger datasets. As such, this book introduces a newly refined methodology of semiotic analysis assisted by Large Language Models. A qualitative approach supplied by semiotics and philosophy of communication is combined with a quantitative approach based on semi-automatic analysis of big social data through Artificial Intelligence, to study the digital propagation of conspiracy theories on Telegram. This method allows for the identification of language patterns that structure conspiracy theories disseminated on online messaging applications, as well as how they relate to the process of radicalisation into extremism. The objective of the refinement and application of computational linguistics tools is not to use machine learning to fully-automatically analyse text, but rather, to serve as a tool for the human analyst. Furthermore, this work approaches the use of AI solutions from a critical point of view, which is to say that computational processes are both methods and objects, being not merely applied, but analysed in themselves. The work detailed in this book was developed within the framework of the Marie Skłodowska-Curie Doctoral Network project ‘VORTEX’ – Coping with Varieties of radicalisation into Terrorism and Extremism (HORIZON-MSCA-2021-DN-01-01).
