Geolibraries

The term geolibrary was introduced in 1998 by the geographer Michael Goodchild to explicitly identify an emerging class of digital system, the “library filled with georeferenced information,” as an important element of the growing global spatial data infrastructure. The geolibrary concept grew out of efforts at organizing and providing shared access to the digital holdings of map [Page 1235]libraries and spatial data repositories. These collections typically include scanned paper maps, satellite imagery, aerial photographs, and geographic information systems (GIS) data files. A key distinction was drawn between this “geographical information,” as representations of the surface and near surface of the Earth, and the vastly broader category of “georeferenced information,” defined as any information referring to or about particular places.

A geolibrary indexes its holdings by location, specified by one or more place names and a geographic “footprint” of any complexity, as well as by the traditional library catalog keys of title, author, and topic. Many digital map libraries and spatial data clearinghouses are then by definition geolibraries, in that their holdings may be browsed and searched by location. It became increasingly apparent that the methods developed for accessing those data could be applied to any digital information objects associated with locations specified by name or footprint. In principle, any library would become a geolibrary if its holdings were comprehensively georeferenced. At its furthest extension, such functionality resembles the Digital Earth vision of systems providing access to all available digitized information related to any specified location.

Geolibrary development—particularly as distributed systems—has raised numerous research challenges and is a valuable context for their solution. These include the following: (a) interoperability for data sharing by means of metadata standards, authority lists, and ontologies; (b) gazetteer development, that is, the association of a particular spatial footprint with potentially numerous place names; (c) efficient spatial indexes for databases and related issues of scale, resolution, and generalization; (d) georeferencing, including the parsing of textual material for named places and related temporal issues; (e) spatiotemporal query algorithms; and (f) interface usability for spatial browsing and search informed by cognitive principles. Significant incremental progress in all these areas has made the prospect of very large, distributed geolibraries increasingly realistic.

In 2004, Goodchild suggested that the next stage of progress in geolibrary development might be from simply delivering information objects (maps, imagery, and other documents) to adding some ability to open and process those objects. In addition to answering the basic query “What do you have about this place?” it could answer many of the type “What is (or was) so interesting about this place?” Such extensions are likely to blur distinctions between the geolibrary and other emerging classes of information systems, including digital atlases and georeferenced digital encyclopedias.