[esip-semanticweb] ok, let's start with GCIS identifiers

[Lynnes, Christopher S. (GSFC-6102)]

On Mar 27, 2012, at 7:32 PM, Curt Tilmes wrote:

> On 03/27/2012 06:11 PM, Lynnes, Christopher S. (GSFC-6102) wrote:
> 
>> Consensus was that we would like to shoot for URIs with RDF content.
>> No objections to going with GCIS naming, so...let's go with it.
> 
> Sounds good.
> 
> BTW, here is some of the data.gov.uk URI info I mentioned:
> 
> http://www.cabinetoffice.gov.uk/sites/default/files/resources/designing-URI-sets-uk-public-sector.pdf
> 
> http://data.gov.uk/resources/uris
> 
> and some still developing W3C GLD suggestions:
> 
> http://www.w3.org/2011/gld/wiki/223_Best_Practices_URI_Construction
> 

So the irony here is that I was led back to a TWC site for the actual specifics on how to construct URIs:  http://logd.tw.rpi.edu/instance-hub-uri-design

GCIS adds some new factors that are not yet accounted for in the TWC recommendation:  (1) GCIS is really a clearinghouse/registry for datasets that are actually held by the agencies, and (2) those agencies use a variety of internal identifiers and (3) identifier use may even vary within an agency.

My suggestion to solve this is:

http://globalchange.gov/id/us/<agency-that-holds-data>/dataset/<id-scheme>/<identifier>

where id-scheme is one of: gcmd, doi or uuid.

An example URI for AIRS Level 3 Daily products would be the following:

http://globalchange.gov/id/us/nasa/dataset/gcmd/GES_DISC_AIRX3STD_V005

An example URI for ORNL DAAC's Global Fire Emissions Database, Version 2 (GFEDv2.1):

http://globalchange.gov/id/us/doe/dataset/doi/10.3334/ORNLDAAC/849

What do you say?

>> Next question is whether service access points (WMS, WCS, OPeNDAP,
>> THREDDS) should also be a property within the GCIS-hosted
>> content. What do you think?
> 
> I think it is reasonable to put them in GCIS to be able to represent
> their relationships with other things.
> 
> Curt

--
Dr. Christopher Lynnes, NASA/GSFC, Code 610.2, 301-614-5185