Impact Estimation of Disinformation Narratives Using Network Causal Inference

• Activity Number: 283 - Statistical Analysis on Social Media Misinformation Campaigns
• Type: Topic-Contributed
• Date/Time: Wednesday, August 11, 2021 : 1:30 PM to 3:20 PM
• Sponsor: Social Statistics Section
• Author(s): Steven Thomas Smith* and Edward Kao and Erika Mackin and Danelle Shah and Olga Simek and Donald B. Rubin
• Companies: MIT Lincoln Laboratory and Lincoln Laboratory at Massachusetts Institute of Technology and MIT Lincoln Laboratory and MIT Lincoln Laboratory and MIT Lincoln Laboratory and Temple University
• Keywords: causal inference; networks; machine learning; social media; disinformation; influence operations

Abstract

This talk presents the causal inference approach used as part of an end to-end framework to automate detection of disinformation narratives, networks, and influential actors. The framework integrates natural language processing, machine learning, graph analytics, and a network causal inference approach. The impact of each account is inferred by its causal contribution to narrative propagation over the network. It accounts for social confounders (e.g., community membership, popularity) and disentangles their effects using an approach based on the network potential outcome framework. Because it is impossible to observe both the realized and the counterfactual outcomes, the missing potential outcomes must be estimated, which is accomplished using a model. We demonstrate this approach's capability on real-world hostile IO campaigns and show that it correctly classifies known IO accounts and networks, and discovers high-impact accounts that escape the lens of traditional impact statistics based on activity counts and network centrality. Results are corroborated with known IO accounts based on US Congressional reports, investigative journalism, and IO datasets provided by Twitter.