Populating the Initial Gazetteer
A relational database was established for the new gazetteer that
provides tables for all of the attributes of the ADL Gazetteer Content
Standard. PDF documents showing the database model are accessible from:
http://www.alexandria.ucsb.edu/
lhill. A document describing the ADL
Gazetteer Content Standard can be found from the ADL Web Site:
http://www.alexandria.ucsb.edu (under Publications/Metadata).
The database currently contains 4840 entries, composed of the following:
US counties: 3111 (bounding boxes)A conversion script to re-categorize the place names from the U.S. Geological Survey (USGS) and the National Imagery and Mapping Agency (NIMA) is ready for testing. This script will allow us to ingest the gazetteer information from these agencies and apply the Feature Type terms from our thesaurus. The script reads the categories present in the incoming data and treats them in one of the following ways: (1) Simple conversion where an existing category directly maps to a new category; (2) Complex conversion where words in the place name are analyzed within a category to decide which new category should be assigned; and (3) General conversion where certain words in place names, no matter what category they are in, result in the assignment of a category.US states: 50 (bounding boxes)
Countries/continents/regions: 171 (bounding boxes)
Volcanoes (Smithsonian Holocene): 1508 (point locations)
We are still in the process of negotiation with the USGS and NIMA to get updated versions of their gazetteers to ingest into the new gazetteer. The advantage of these gazetteers is that they are extensive (approximately 6 million entries in total); the disadvantage is that they contain point locations only.
We are also in the process of talking to other potential sources of gazetteer entries with bounding box extents and expect to be able to add substantially to the new gazetteer through time.