Sign up for our weekly newsletter!
Title: Mining for Conspiracies: A pipeline for the discovery of conspiracy theories on social media
Speaker: Tim Tangherlini is a Professor in the Dept of Scandinavian here at UC Berkeley.
Abstract: In this brief presentation, I explain some of the challenges in developing a largely automated machine learning pipeline for working with social media posts. The goal of our work is to discover the underlying generative narrative frameworks for conversations in social media. Initially developed to make sense of the diffuse discussions about vaccine hesitancy and exemption seeking on parenting blogs, we refined our methods over the past five years to allow us to work with large, noisy social media data sets. Recently, we have use our approach to understand the structural differences between conspiracies and conspiracy theories, focusing on Bridgegate and Pizzagate. More recent work has focused on the totalizing QAnon conspiracy theory, and the complex storytelling environment of the Covid-19 pandemic. In this brief talk, I will explain aspects of the pipeline and present some of the results of our work.
Check out future working group sessions.
