This project develops a new body of spatial analytic methods for data intensive research in the social and behavioral sciences. Through an interdisciplinary team of geographic information scientists, public policy scholars, and computer scientists, this project provides researchers with the next generation of scalable, high performance spatial analytical techniques. The research targets existing challenges that have hindered the more extensive and intensive utilization of geospatial data from the US Census, American Community Survey, and social media through a combination of new computational algorithms for regionalization, new approaches to computational inference, and use of scalable spatial and spatiotemporal data management techniques. Research advances from this project will be integrated into new courses in data science, computer science, and public policy. Project results and materials will be disseminated in several ways including open source spatial analysis software repositories, tutorials and workshops for social scientists, and publications in scientific outlets.
The project's goals are to enhance the open source spatial analysis library PySAL in four main directions. The first innovation concerns new methods for the treatment of uncertainty in the American Community Survey data. Second, a new framework for statistical inference in the space-time analysis of segregation will be developed. Third, scalable algorithms for regionalization and spatially explicit aggregation of areal units will be produced. Finally, the project will develop a new scalable spatio-temporal data management layer to support the application of these three sets of analytics to large datasets. The software developed here will be included in enhanced (and new) modules in PySAL. This will afford the development of new types of applications using different interfaces from plug-ins for desktop GIS, both open source and proprietary, to the development of software as service systems and the ability of installations where on-premise requirements are binding (Census Research Data Centers). This portfolio of use cases will support a wide range of scientific disciplines and will provide direct benefits to society through new geospatial tools that will improve the state of the art in urban planning, economic development, and spatial public policy.
This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.
