Spatial Arrangement as a part of Geospatial Feature Ontologies

SPATIAL ARRANGEMENT AS A PART OF GEOSPATIAL FEATURE ONTOLOGIES Jonathan R. Clark, Ph.D George Mason University, 2012 Dissertation Director: Dr. Anthony Stefanidis The geospatial sciences have always employed ontologies to describe spatial data, conduct analysis, and to design the systems and methods for conducting geospatial operations. While research on geospatial ontologies has addressed non-spatial aspects of features on the terrain, space and spatial relationships are usually treated as separate from the nature of geospatial objects. This work presents an approach for making spatial context an integral part of geospatial feature ontologies. Specifically, the spatial arrangement of the components of a complex feature class (features composed of several simple features) was examined as a potential “spatial signature” of that feature class. This work suggests an approach for including spatial relationships and metrics in a prototypical feature ontology, and then using these metrics to judge the similarity of metrics from other feature classes to that prototype. Container terminals, a type of commercial maritime terminal, were used as a test case with a machine-readable ontology developed using spatial metrics from sample terminals world-wide. Like metrics were also collected for test cases representing several types of facilities, and compared to the prototype ontology. Using a simple similarity model, comparison of the test cases to the prototype resulted in partial success in judging similarity of the test cases to that prototype. Variations in facilities layouts and the simple logic of the similarity model were the likely causes of reduced success in similarity judgments.