Discovering spatial relationships in geographic information databases for geodatasets integration

Abstract

Describes a framework for developing the integration of multiple single theme datasets (maps). The novelty of this approach is that it is based on spatial data mining, in particular on spatial association rules. A spatial association rule is a rule indicating certain association relationships among of a set of spatial and possibly some non-spatial predicates. Prior to the generation of the association rules, the source maps are represented in a common vector structure. Spatial hierarchies are defined to determine the characteristics of the features belonging to distinct classes. A distance restriction to limit the search space is specified as a part of the user query. Structure, levels of abstraction and reduction of the space guarantee the use of only the necessary data to answer the user query. A characterization of the uncertainty in the fused map is given as the measure of correction of the framework.