deprecate/map UNCEFACT geospatial stuff in favor of OGC GeoSPARQL about spec-jsonld HOT 6 CLOSED

Please note that GeoSPARQL 1.1 now contains a better explanation of the Simple Features geometry types classes (https://opengeospatial.github.io/ogc-geosparql/geosparql11/spec.html#_geometry_class_hierarchy) and an annex listing alignments with other ontologies (https://opengeospatial.github.io/ogc-geosparql/geosparql11/spec.html#_annex_e_alignments_informative).

For the classes listed above, I suspect these mapping might be useful:

Assuming that there is nothing really different between a 'Point' and a 'GeographicalPoint' (I don't know UNCEFACT) then:

I can find no mappings for the following

NOTE:* having SF classes in GeoSPARQL doesn't do anything for you. GeoSPARQL functions operate on geometry literals, not RDF classes, so, if you want to calculate intersections with a GeoSPARQL engine, you need to make polygon, point etc literals like this:

:feature-x
    geo:hasGeometry [
        geo:asWKT "POLYGON ((...))"^^geo:wktLiteral ;
   ] ;
.

:feature-y
    geo:hasGeometry [
        geo:asWKT "POINT (...)"^^geo:wktLiteral ;
   ] ;
.

WKT, GML etc.

from spec-jsonld.

As @nissimsan says, we can only deprecate not delete, eg

uncefact:Circle a rdfs:Class; owl:deprecated true;
  skos:scopeNote "Deprecated in favor of OGC GeoSPARQL geometries".

We should also map, eg

uncefact:Polygon a rdfs:Class; owl:deprecated true;
  skos:scopeNote "Deprecated in favor of OGC GeoSPARQL geometries";
  dct:isReplacedBy sf:Polygon.

The new draft spec is https://opengeospatial.github.io/ogc-geosparql/geosparql11/spec.html and adds serializations GeoJSON, KML, DGGS but it will be some time before implementations pick up.

@nicholascar, @dr-shorthair could you review my prev comments and help with mapping?

from spec-jsonld.

@VladimirAlexiev, we are discussing this on the call now.
It is not the role of this project to interfere with the semantics. However, the two actions we will take are:

• List equivalence vocabularies for overlapping areas.
• Pass back this recommendation to UN modelling teams suggesting removal of this part of the BSP model.

from spec-jsonld.

This is an issue for the UNCEFACT business domains / vocabulary management. I suggest
y

• we start a list of alternative vocabularies (eg schema.org) - not stating preference just listing them
• we feeedback to the business domain that we suggest they may like to deprecate the UN geospatial vocabulary in favour of another.

from spec-jsonld.

@nicholascar, awesome! This is a great format for this NDR section. I love it - thank you!

from spec-jsonld.

@nicholascar , I added a section about this now.
#75

from spec-jsonld.