DISCOVERING STRUCTURE IN GEOGRAPHICAL METADATA
Igor T. Podolak · 2004
Metadata are data about data. Geographical metadata describe geospatial data: maps, satellite images or other geographically referenced material. The two characteristics of geographical metadata, high dimensionality and diversity of attribute data types, present a problem for automatic data mining. We present an approach for exploration of geographical metadata which is based on an integration of visual data mining and an automatic data mining method, clustering. Clustering discovers hierarchical structure in the metadata in order to help the user with the exploration. A special visualisation of the structure is integrated in the visual data mining concept.