An automated pipeline for the discovery of conspiracy and conspiracy theory narrative frameworks: Bridgegate, Pizzagate and storytelling on the web

Abstract

Although a great deal of attention has been paid to how conspiracy theories circulate on social media, and the deleterious effect that they, and their factual counterpart conspiracies, have on political institutions, there has been little computational work done on describing their narrative structures. Predicating our work on narrative theory, we present an automated pipeline for the discovery and description of the generative narrative frameworks of conspiracy theories that circulate on social media, and actual conspiracies reported in the news media. We base this work on two separate comprehensive repositories of blog posts and news articles describing the well-known conspiracy theory Pizzagate from 2016, and the New Jersey political conspiracy Bridgegate from 2013. Inspired by the qualitative narrative theory of Greimas, we formulate a graphical generative machine learning model where nodes represent actors/actants, and multi-edges and self-loops among nodes capture context-specific relationships. Posts and news items are viewed as samples of subgraphs of the hidden narrative framework network. The problem of reconstructing the underlying narrative structure is then posed as a latent model estimation problem. To derive the narrative frameworks in our target corpora, we automatically extract and aggregate the actants (people, places, objects) and their relationships from the posts and articles. We capture context specific actants and interactant relationships by developing a system of supernodes and subnodes. We use these to construct an actant-relationship network, which constitutes the underlying generative narrative framework for each of the corpora. We show how the Pizzagate framework relies on the conspiracy theorists' interpretation of "hidden knowledge" to link otherwise unlinked domains of human interaction, and hypothesize that this multidomain focus is an important feature of conspiracy theories. We contrast this to the single domain focus of an actual conspiracy. While Pizzagate relies on the alignment of multiple domains, Bridgegate remains firmly rooted in the single domain of New Jersey politics. We hypothesize that the narrative framework of a conspiracy theory might stabilize quickly in contrast to the narrative framework of an actual conspiracy, which might develop more

Conspiracy in the Time of Corona: Automatic detection of Emerging Covid-19 Conspiracy Theories in Social Media and the News

“…In prior work [67], we have shown that the final network comprising subnodes and their relationship edges is not sensitive to the exact size and content of the contextual groups derived at this stage. This CG grouping, which we undertake prior to applying k-Means clustering on word embedding, enables us to distill the noun phrases into groups of phrases that have a semantic bias.…”

“…We model narratives as generated by an underlying graphical model [67]. The narrative graph is characterized by a set of n nodes representing the actants, a set of r relationships R = {R 1 , R 2 , .…”

“…Inspired by the narratological work of Algirdas Greimas [28], and the social discourse work of Joshua Waletzky and William Labov [36], we devise an automated pipeline that determines the narrative frameworks that form part of the bedrock of diverse knowledge domains, in this case those related to the Covid-19 pandemic [67]. We also borrow from George Boole's famous definition of a domain of discourse, recognizing that in any such domain there are informal and constantly negotiated limits on what can be said [10].…”

“…We also borrow from George Boole's famous definition of a domain of discourse, recognizing that in any such domain there are informal and constantly negotiated limits on what can be said [10]. We conceptualize a narrative framework as a network comprising the actants (people, organizations, places and things) and the interactant relationships that are expressed in any storytelling related to the pandemic, be it a journalistic account or an informal anecdote [69] [67]. In this model of story telling, individuals usually activate only a small subset of the available actants and interactant relationships that exist in a discourse domain, thereby recognizing that individual storytelling events are often incomplete.…”

“…By applying the narrative framework discovery pipeline to tens of thousands of social media posts and news stories, all centered on conspiracy theories related to the Covid-19 pandemic, we uncover five central phenomena: (i) the attempt by some conspiracy theorists to incorporate the pandemic into well-known conspiracy theories, such as Q-Anon; (ii) the emergence of new conspiracy theories, such as one suggesting that the 5G cellular network is the root cause of the virus aligning the domains of telecommunications, public health, and global trade; (iii) the alignment of various conspiracy theories to form larger ones, such as one suggesting that Bill Gates is using the virus as a cover for his desire to create a worldwide surveillance state through the enforcement of a global vaccination program, thereby aligning it with anti-vaccination conspiracy theories and other conspiracy theories related to global cabals; (iv) the nucleation of potential conspiracy theories, such as #filmyourhospital, that may grow into a larger theory or be subsumed in one of the existing or emerging theories; and (v) the interaction of these conspiracy theories with the news, where certain factual events, such as the setting up of tents in Central Park for a field hospital to treat the overflow of patients, are linked to conspiracy theories. In this particular case, the tents of the field hospital are linked to central aspects of the Pizzagate conspiracy theory, specifically child sex-trafficking, underground tunnels, and the involvement of highly visible public figures [34] [67].…”