Mapping Conspiracies Narratives about the Great Reset: A Computational Text Analysis Approach

In the realm of Text Analysis and Natural Language Processing, an often-employed strategy involves the utilization of unsupervised learning methods, such as Topic Modeling, to reduce texts into to their potential main themes. In this research, we employ a different approach, using network analysis of text components. Roberto Franzosi (2010) pioneered narrative analysis within sociology by employing the reduction journal text to triplets - who made what (violence) to whom? - to examine the emergence of fascism in the 1930s in Italy. We construct our dataset by gathering texts by conspiracy theorists, such as Alex Jones and Alexandr Dugin, with the aim of comprehending their divergent framings, charting their primary arguments and motifs of conspiracist perspectives on the Great Reset. The Great Reset was a proposal by the World Economic Forum during the COVID-19 pandemic that aimed for changes in the global economy, which quickly garnered attention in conspiracist circles as evidence of a globalist/communist attempt to establish a dictatorial regime marked by numerous restrictions on freedom, food, abolition of private property, through the implementation of social credits and digital currencies. We begin extracting the grammar components using Part-of-Speech tagging, and then extracting the semantic relations to capture semantic roles, like object-action(verb)-subject. Using embeddings into specific types of words, we then clustered the extracted graphs into common themes, and filtered some phrases to retain the more meaningful ones. Dugin focus in the West/Western and in the "spontaneous response of the human masses" where Russia would play a vital role, while Alex Jones focus on Harari and how Trilateral Commission guide democrats presidents to impose a global government. Although some claims may be outdated, even among conspiracists, we believe that many of these claims and the bigger picture linking them, survive today.