Re: Property "geographic identifier" in LOCN (was: Re: ISA Core Location Vocabulary)

Hi,
In the RDF world, every resource is identified by a URI.

In the GIS world, a geographic feature is usually identified by a
geographic identifier.

In the US for example, "The Geographic Names Information System Identifier
(GNIS ID)
is a variable length, permanent, numeric identifier of up to ten digits in
length that
identifies each entity uniquely within the nation.  The GNIS is the new
American National
Standards Institute (ANSI) national standard code for several entity
types.  Because each
entity's GNIS ID is permanent, it will not change if the entity changes its
name or if
creation of a new entity changes the alphabetic sort. [...]" [1].

My opinion is that the vocabulary must contain a property like the proposed
"geographic identifier" to ensure interoperability with existing geospatial
tools/datasets. We have found this property to be of great importance, as
far as
interoperability and adoption of Linked Data technologies from domain
experts are
concerned. In the past, we have developed a real-time wildfire monitoring
service [2]
for the National Observatory of Athens, where we used satellite data along
with
Linked Geospatial Data to produce fire maps. At any given point, we were
able to
export the result of a stSPARQL/GeoSPARQL query as an ESRI shape file (that
included geographic identifiers), that allowed the domain experts to use
the (enriched
with Linked Geospatial Data) shape files with their existing (I would not
call them
legacy) tools/datasets/processing chains/infrastructure in general. This
allowed them
to load the shape files to ArcGIS for example, do a thematic join (based on
the
geographic identifier that was preserved) with auxiliary datasets that they
are using
in a daily basis and are not published as Linked Open Data and produce
their final
products.

After reading the description of the geographic identifier property, I
think that this is
property is used as spatial version of owl:sameAs between identifiers
defined by
different authorities/publishers. In addition, a publisher can use this
property to
assert that a location has a specific identifier (e.g., a URN) which is
very useful in
practice in order to ensure interoperability with existing infrastructures.

Best,
Kostis

[1] http://www.census.gov/geo/reference/gtc/gtc_gnisid.html
[2] http://test.strabon.di.uoa.gr/


===================================================
Kostis E. Kyzirakos, Ph.D.
Centrum voor Wiskunde en Informatica
DB Architectures (DA)
Office L320
Science Park 123
1098 XG Amsterdam  (NL)
tel: +31 (20) 592-4039
mobile: +31 (0) 6422-95345
e-mail:  kostis@cwi.nl
===================================================


On Thu, Jan 2, 2014 at 9:23 AM, Raphaël Troncy <raphael.troncy@eurecom.fr>wrote:

> Dear all,
>
>
>  Is it possible to give an RDF example of meaningful usage of "geographic
>> identifier"? I think that could help  in understanding the issue.
>>
>
> I find myself with the same opinion than Frans. I don't yet understand the
> purpose of this "geographic identifier" property, although I understand the
> need for a vocab to have its own way to describe its specific identifiers
> if there are some constraints on how those identifiers should be
> interpreted. When this property was introduced in the vocab? Is it a
> mapping with a specific field from the INSPIRE schemas?
> Best regards.
>
>   Raphaël
>
> --
> Raphaël Troncy
> EURECOM, Campus SophiaTech
> Multimedia Communications Department
> 450 route des Chappes, 06410 Biot, France.
> e-mail: raphael.troncy@eurecom.fr & raphael.troncy@gmail.com
> Tel: +33 (0)4 - 9300 8242
> Fax: +33 (0)4 - 9000 8200
> Web: http://www.eurecom.fr/~troncy/
>
>