Using Formal Ontology for Integrated Spatial Data Mining

Abstract

With increasingly available amount of data on a geographic space, spatial data mining has attracted much attention in a geographic information system (GIS). In contrast to the prevalent research efforts of developing new algorithms, there has been a lack of effort to re-use existing algorithms for varying domain and task. Researchers have not been quite attentive to controlling factors that guide the modification of algorithms suited to differing problems. In this study, ontology is examined as a means to customize algorithms for different purposes. We also propose the conceptual framework for a spatial data mining (system) driven by formal ontology. The case study demonstrated that formal ontology enabled algorithms to reflect concepts implicit in domain, and to adapt to users' view, not to mention unburdened efforts to develop new algorithms repetitively. © Springer-Verlag 2004.