SPSS Statistics for Data Analysis and Visualization

Chapter 8 Geospatial Analytics

The majority of data that businesses deal with is associated with a particular location. For many problems, incorporating where events occur can improve predictions and help you better understand the data. Almost all spatial data exhibits some type of spatial dependence, and exploiting that dependence is what spatial analysis is all about.

Two new spatial analytic techniques have been added to SPSS Statistics Ver-sion 23: Geospatial Association Rules and Spatio-Temporal Prediction. Both techniques are available to all users. Both techniques can take a set of data that has an associated set of X and Y coordinates, or has a field that associates it with a particular map file (for example, a ZIP code) and then generate predictions for an area given a set of map locations and additional variables associated with those areas.

This chapter provides a walkthrough of each technique applied to different case studies of crime analysis. Crime tends to be highly clustered in certain areas, and so many police departments conduct geographic analysis to identify these hot spots of crime. Police departments then frequently deploy more resources at these hot spots, such as conducting more patrols, writing more traffic tickets, or having community police officers meet with community members to discuss local problems and potential remedies. Each of the new geospatial analytic techniques in SPSS can facilitate the types of analyses many police departments or other ...