Spatiotemporal Model Fusion: Multiscale Modeling of Civil Unrest

• Activity Number: 686
• Type: Contributed
• Date/Time: Thursday, August 13, 2015 : 10:30 AM to 12:20 PM
• Sponsor: Section on Bayesian Statistical Science
• Author(s): Andrew Hoegh* and Marco A.R. Ferreira and Scotland Leman
• Companies: Virginia Tech and Virginia Tech and Virginia Tech
• Keywords: Sequential Monte Carlo ; Particle Learning ; Data Fusion

Abstract

Civil unrest is a complicated, multifaceted social phenomenon for which forecasting of upcoming protests is a challenging problem. Relevant data for predicting future protests consist of a massive set of heterogenous data sources, primarily from social media. Using a modular approach to extract pertinent information from disparate data sources, a spatiotemporal multiscale framework is developed to fuse predictions from algorithms mining social media. This novel multiscale spatiotemporal model is developed to satisfy four essential requirements: (1) be scalable to handle massive spatiotemporal datasets, (2) incorporate hierarchical predictions, (3) accommodate predictions of differing quality and uncertainty, and (4) be flexible, allowing revisions to existing algorithms and the addition of new algorithms. This talk details the challenges posed by these four requirements and outlines the benefits of our novel multiscale spatiotemporal model relative to existing methods.