Lab alum Sean Fitzhugh, along with lab member Ben Gibson, lab alum Emma Spiro, and lab PI Carter Butts, have published their article “Spatio-temporal Filtering Techniques for the Detection of Disaster-related Communication” in an upcoming special issue of Social Science Research on big data. This paper develops a family of techniques for filtering large streams of communication across time and space in order to detect signals of hazard-related rumoring. They reliably detect signal of rumoring activity across a variety of disasters affecting a wide range of populations, from dense urban areas to sparse rural areas. The results highlight spatio-temporal variation of rumoring activity in response to impending and realized disasters. Additionally, they illustrate a case study of how message content varies across time and space during a disaster.
The article can be found online here.