- From: Krzysztof Janowicz <janowicz@ucsb.edu>
- Date: Tue, 19 May 2015 23:17:04 -0700
- To: <Simon.Cox@csiro.au>, <L.Svensson@dnb.de>, <raphael.troncy@eurecom.fr>, <Kerry.Taylor@csiro.au>, <eparsons@google.com>
- CC: <public-sdw-wg@w3.org>
I agree with Simon here. There will always be Linked Data 'leaf nodes' that will not (and do not have to) be triplified. If I recall correctly, the GeoSPARQL group had similar discussions. In almost all cases (I can think of), for instance, having a full RDF serialization of a complex polyline feature does not add any value (compared to WKT). This is even not about Linked Data versus Semantic Web reasoning, it is simply about the added value (or the lack of it). Best, Krzysztof On 05/19/2015 10:30 PM, Simon.Cox@csiro.au wrote: > Raphaël: how is 'semanticizing' the description of CRS helpful? As Peter and I have shown there are existing XML-based services that deliver the entire EPSG CRS dataset in fully structured form (which covers the lat/lon vs lon/lat issue Lars). Given that these services have reliable URIs (based on the EPSG identifiers), contain links (to the component elements like CS, Datum, Axis, etc), and are in an open format (GML/XML), we are already up to about 4 1/2-star linked data. > > I'm a big fan of RDF and OWL, partly because of the scalability and flexibility, and tool support. But there are some boundaries over which the value add of RDF is vanishingly small, particularly if some 'linked data' that is already available. I question whether effort is wisely spent here, compared with some other parts of the puzzle which are much less evolved right now. > > Kerry raised the issue of scope, and suggested that the goal should be n-star linked spatial data. I agree, but we need to be clear that "linked data" != "semantic web with full reasoning", so need to be careful about balance here. > > -----Original Message----- > From: Svensson, Lars [mailto:L.Svensson@dnb.de] > Sent: Tuesday, 19 May 2015 9:51 PM > To: Raphaël Troncy; Cox, Simon (L&W, Highett); Taylor, Kerry (Digital, Acton); eparsons@google.com; janowicz@ucsb.edu > Cc: public-sdw-wg@w3.org > Subject: RE: UCR issue: phrasing of CRS requirement(s) > > On Tuesday, May 19, 2015 10:09 AM Raphaël Troncy wrote: > >>> Thanks Kerry - that's essentially the way I see it, if by "linked >>> data representation" you are implying RDF. I would like to ask those >>> people advocating a new CRS encoding in RDF, what this would be >>> useful for? >> Well of course, this is a very "niche" usage, but typical use cases >> are for getting an explicit semantic description of how a CRS has been >> built so that you can, for example, query for all CRSs that use a >> specific Datum, or, more simply, ask for the EPSG identifier >> corresponding to the URI of a CRS, etc. > Another case would be to get information about lat/long vs. long/lat. > > Best, > > Lars -- Krzysztof Janowicz Geography Department, University of California, Santa Barbara 4830 Ellison Hall, Santa Barbara, CA 93106-4060 Email: jano@geog.ucsb.edu Webpage: http://geog.ucsb.edu/~jano/ Semantic Web Journal: http://www.semantic-web-journal.net
Received on Wednesday, 20 May 2015 06:17:42 UTC