Скачать или смотреть Ontology as Knowledge Base for Spatial Data Harmonization

Base on experience from many international project (e.g. Plan4all, SDI-EDU, Habitats or Humboldt) we has learned that many users are limited during processing spatial data sets from different sources. These data sets have usually various properties such as reference systems, data models, data formats and theme. The common user has some common conception of the result but they are not able to define any requirements on process of transformation of source and target data. The composition of harmonization process (as a sequence of simple step to limit some aspects of spatial data heterogeneity) represents one of the important task of current geoinformatics and geomatics.

To improve a communication between spatial data harmonization experts and common users (state administrative officers, spatial planners, teachers...) the expert system could be established. This systems should be based upon a knowledge base that will formalize information and knowledge appropriate to data harmonization and sharing.
The goal of this paper is to introduce a domain ontology as the fundamental part of knowledge base. The ontology describes the data sets, their layers and properties (e.g. reference system, way of storage, theme or geometry). After reasoning original ontological structure (with explicit information) the inferred ontology is created. This ontology contain two key class -- Layer (the data sets are grouped on the basis of theme) and Harmonization (particular layers are assigned to various harmonization sub-steps). The system uses the GEMET Thesaurus as a source of keywords defining the data themes.

The article Ontology as Knowledge Base for Spatial Data Harmonization is divided into four parts (except Introduction). The section "Theoretical Background" summarizes the approaches of spatial data harmonization as well as theory of domain ontologies (principles, methodology). The next part describes developed ontology. Then the sub-chapter focused on particular examples of data harmonization with using ontologies follows. The conclusion summarizes and discusses the results and proposes next steps of further development.

by Otakar Cerba, Karel Charvat
University of West Bohemia, Plzen, Czech Republic
Czech Center for Science and Society, Praha, Czech Republic