Data annotation question

11 views
Skip to first unread message

Bernhard Haslhofer

unread,
Feb 27, 2012, 9:11:43 PM2/27/12
to oac-discuss
Hi Rob et al,

I am currently specifying how we are going to represent georeferences on maps using OAC. A georeference
basically associates a place with a point (x,y) on a map.

We have two different scenarios:
1.) the place is identified by a dereferencable HTTP URI, which delivers RDF
2.) we need to embed the lat/long coordinates directly in the annotation RDF document, because no place URI is available.


Here is how I would model the first case:


An alternative could be to skip the oac:GeoReference typing and introduce a property ex:geoReferences between the body and the target. Both have pros and cons. A pro is probably that you can easily fetch all Georeferences by SELECT ?x WHERE ?x a oac:GeoReference.




And here is the solution we currently have on the table for the second case:

    a oac:GeoReference ;
    oac:hasBody <urn:uuid:926a3b20-617c-11e1-b86c-0800200c9a66> ;

<urn:uuid:926a3b20-617c-11e1-b86c-0800200c9a66>
    gn:name "City of Ithaca"^^<http://www.w3.org/2001/XMLSchema#string> ;
    wgs84_pos:lat "43.29225"^^<http://www.w3.org/2001/XMLSchema#float> ;
    wgs84_pos:long "-84.60472"^^<http://www.w3.org/2001/XMLSchema#float> .


Is this the way we should model it or should we really serialize the <urn:uuid…> statements to a string, as the current spec proposes. I believe that the drawback of this is that important data (lat/long) become unprocessable by clients unless the deserialize the string and re-build the model; which is not very practical.

Thanks for advice / opinions / etc.

Bernhard


-----------------------------------------------------
Bernhard Haslhofer
Postdoc Associate
Cornell Information Science
301 College Ave. 
Ithaca, NY 14850
WWW: http://www.cs.cornell.edu/~bh392/
Skype: bernhard.haslhofer

Reply all
Reply to author
Forward
0 new messages