Conspiracy to Commit: Information Pollution, Artificial Intelligence, and Real-World Hate Crime

Abstract

Is demand for conspiracy theories online linked to real-world hate crimes? By analyzing online search trends for 36 racially and politically-charged conspiracy theories in Michigan (2015-2019), we employ a one-dimensional convolutional neural network (1D-CNN) to predict hate crime occurrences offline. A subset of theories including the Rothschilds family, Q-Anon, and The Great Replacement improves prediction accuracy, with effects emerging two to three weeks after fluctuations in searches. However, most theories showed no clear connection to offline hate crimes. Aligning with neutralization and differential association theories, our findings provide a partial empirical link between specific racially charged conspiracy theories and real-world violence. Just as well, this study underscores the potential for machine learning to be used in identifying harmful online patterns and advancing social science research.

References in corpus (8)

• An Empirical Evaluation of Generic Convolutional and Recurrent Networks for Sequence Modeling
• Deep learning for time series classification: a review
• Deep Learning using Rectified Linear Units (ReLU)
• InceptionTime: Finding AlexNet for Time Series Classification
• Exploring the Effects of COVID-19 Containment Policies on Crime: An Empirical Analysis of the Short-term Aftermath in Los Angeles
• Permutation-based Causal Inference Algorithms with Interventions
• Explainable Machine Learning for Predicting Homicide Clearance in the United States
• Robust inference of causality in high-dimensional dynamical processes from the Information Imbalance of distance ranks