This is a potential GSoC project - I'd see it firstly being a separate
codebase as it develops, then get some people using it and providing
feedback and, if it's popular, it can move into Jena proper. The nice
thing is that doesn't (err "shouldn't") require modifications to the
main modules as it should use standard extension mechanisms. However,
if the mechanisms prove inadequate, we can fix the extension mechanisms.
This is the very similar to the status of jena-text - it's in
/Experimental so anyone can look, use and contribute but it's not in the
Jena distribution. If and when it's considered stable, it could become
part of trunk. That's a group decision to be made.
The project is also open ended depending on how "simple" it is. There
are some jumps in going from, say, points to polygons in the data that I
think are just too big for the GSoC timeframe but there are potential
additional query functions to just "near".
At least look at GeoSPARQL [1] to see what a complete solution looks
like. I think that the sweet spot is something simpler (and less
capable) because the average web developer, even if writing SPARQL,
isn't looking for a complete solution to all geospatial use cases. They
are looking for something easier (= smaller). There is space for both
approaches.
Being able to use Lucene-spatial for the index, and have it instep with
the data, rather than an external rpcoess like PostGIS (c.f. Parliament)
makes it much more usable.
It also avoids some licensing issues - some geo libraries are GPL,LGPL
which makes them unsuitable for Apache.
Combine this with any work on a better Fuseki UI, and we have a nice,
easy-to-use SPARQL server that gets people up and running with
interesting features beyond basic RDF storage and query.
Andy
[1] http://www.opengeospatial.org/standards/geosparql
On 15/04/13 22:52, Gyanasekaran Radhakrishnan wrote:
This is an awesome idea which could be extensively used if implemented (I
would!). Is this going to be a stand-alone project for GSoC? I would be
interested in working in this.
Thanks
~Gyan
On Mon, Apr 15, 2013 at 9:37 AM, Andy Seaborne <a...@apache.org> wrote:
=== Spatial query
JENA-10
Not all spatial queries are complicated. Many use case are covered by
provision of a single facility like get all objects within a given radius
or within a given bounding box. GeoSPARQL which is a complete approach to
geospatial query but it is complicated and driected more towards the
specialist, not the average linked data developer with SPARQL knowledge.
This project is to provide a one or two basic geospatial query functions.
Example: return places within 10 kilometers of Bristol UK (which as
latitude/longitude of 51.46,2.6).
SELECT ?placeName
{
?place spatial:query (51.46 2.6 10) .
?place rdfs:label ?placeName
}
We have a similar property function architecture in jena-text [0] so this
project is to take that concept and apply it to geospatial information.
Lucene spatial could be used for the spatial index. There is a simple
vocabulary for expressing WGS80 information [1]; there is also the point
information in WKT [2].
This project is not primarily about geospatial processing; it is concerned
with the indexing and querying simple, existing geospatial data and
integration with Fuseki.
[0]
http://jena.staging.apache.**org/documentation/query/text-**query.html<http://jena.staging.apache.org/documentation/query/text-query.html>
[1] http://www.w3.org/2003/01/geo/
[2]
http://en.wikipedia.org/wiki/**Well-known_text<http://en.wikipedia.org/wiki/Well-known_text>
