From the examples you've given how about:
1. Start with the first (most detailed) element in the hieararchy.
2. Moving up the hieararchy, add on the first "inhabited place" found,
if any.
3. Continuing to move up the hieararchy, add on the first "nation"
found, if any.
On 9/17/2012 3:12 PM, [email protected] wrote:
We use the Getty Thesaurus of Geographic Names for coding place names in our
museum and archival cataloguing systems. We're currently struggling with the
best way to display and make these terms searchable in our online database.
Currently we're just displaying the term itself, which is flawed, because just seeing
"Springfield" or "Florence" doesn't give the user enough information to figure
out where something was really made.
But we're finding that the number of variant place types in TGN makes it hard
to figure out a concise way of indiciating a more detailed place name that will
work consistently across all entries in the thesaurus.
For example, the full hierarchy for Florence (the one in Italy) is
Florence (inhabited place), Firenze (province), Tuscany (region), Italy
(nation), Europe (continent), World (facet)
Neigborhoods and other local subdivisions can be even more of a mouthfull:
Notting Hill (neighborhood), Kensington and Chelsea (borough), London
(inhabited place), Greater London (metropolitan area), England (country),
United Kindom (nation), Europe (continent), World (facet)
Ideally I'd probably like to show the above as "Florence, Italy" and "Notting Hill,
London, England"
But I'm having trouble coming up with an algorithm that can consistently spit
these out in the form we'd want to display given the data available in TGN.
Would welcome any ideas or feedback on this.
Thanks,
David
__________
David Dwiggins
Systems Librarian/Archivist, Historic New England
141 Cambridge Street, Boston, MA 02114
(617) 994-5948
[email protected]
http://www.historicnewengland.org
