Indexing and Bowsing Digital Maps with Intelligent Thumbnails

With the increasing number of digital maps and other geo-referenced data that are available over the Internet, there is a growing need for methods to preview the content of such data sources, and to evaluate their relevance with respect to complex spatial and thematic search criteria. Analogous to the digital indices for full-text searches on text documents, we introduce highly condensed, machinereadable indices of digital maps. The purpose of these “intelligent thumbnails” is to support sophisticated queries of the type concept@location. An intelligent thumbnail is based on a projection of the thematic content of a digital map onto a standard reference tessellation. To make the thumbnail highly exchangeable in an distributed and heterogeneous environment, the underlying standard reference tessellation (SRT) is qualitatively abstracted from a polygonal representation to a connection graph. In combination with a hierarchical placename structure and a concept ontology, it is used to evaluate the spatial and thematic relevance of the indexed data sources with respect to spatio-thematic queries.