Information Mediation Across Heterogeneous Government Spatial Data Sources

Spatial information is a common component of databases maintained by governments at different hierarchical levels, from local to federal. Even within a single municipality, information about particular spatial objects may differ in measurement and representation type, content of attributes, data quality, amount and structure of metadata, currency, etc. Such heterogeneity becomes a serious obstacle in applications that require querying across multiple spatial databases. To address this problem, we extend the principles of conventional information mediation to processing of spatial queries against multiple heterogeneous spatial data sources and services. The spatial mediation infrastructure we are developing, includes a set of spatial wrappers which expose capabilities of spatial data sources and services in a homogeneous way, spatial mediators for query planning and orchestrating query execution, and spatial presentation services. This paper focuses on the specifics of spatial information mediation, presenting a model of source spatial capabilities for mediation. Using a running example, we demonstrate generation of feasible query evaluation plans based on source spatial coverage declaration and function signatures for all supported spatial data interchange and spatial data transformation functions.