Geovisual Analytics

Geospatial analysts increasingly are finding that they must deal with very large data sets (Big Data). Interpreting Big Data requires combining our human visual processing capabilities with the analytical capabilities of computers. The result is the rapidly growing field of visual analytics, which is termed geovisual analytics when geospatial data are used. The chapter begins by considering the characteristics and limitations of Big Data, and then stresses the origin and characteristics of geovisual analytics, stressing the capability to “make sense” out of multivariate spatiotemporal data and to “discover the unexpected.” This is often accomplished using Ben Shneiderman&s;s information-seeking mantra. Next, the self-organizing map (SOM) is introduced, which can be used for both clustering and dimension-reduction. The bulk of the chapter illustrates six examples of geovisual analytics: (1) TaxiVis, a system for visualizing taxi trips in New York City; (2) the use of mosaic diagrams for visualizing spatiotemporal data; (3) CarSenToGram, an approach for visualizing Twitter data; (4) Crowd Lens, a tool for visualizing OpenStreetMap contributions; (5) the use of a SOM for sense-of-place analysis; and (6) the Australian Cancer Atlas. A table of 25 additional examples is utilized to illustrate the broad range of geovisual analytics applications. Given the increasing importance of dashboard interfaces in geovisual analytics, the chapter concludes with a summary of resources for developing and evaluating dashboards.

Computer Science · Public Relations and Crisis Communication