Re: Semantic Bleachers: was Re: Afghanistan War Diary as topic map in Maiana

In message <1843333554.20101101205712-0vbsL0CoYrBBDgjK7y7TUQ@public.gmane.org>, Lars Heuer 
<[email protected]> writes

>> What I have been trying to ask is: "What steps are those?"
>
>I don't know. I want to query Wikipedia and maybe I want to update
>Wikipedia through TMQL in the future. I want to stay in my preferred
>technology. We can also translate TMQL to SPARQL, though.

I think supporting updating would be a challenge.  Is there a facility 
for updating Wikipedia pages from the DBpedia environment using SPARQL 
updates?  If so, we could piggy-back onto that.

There is also the question of the relative expressivity of SPARQL vs. 
TMQL.  What query possibilities would be lost by having to virtualise 
TMQL queries as SPARQL?

I like the idea of "virtual topic maps", but I think we need a 
justification for this approach which goes beyond staying in a preferred 
technology. (Others might disagree!)  As has been discussed recently, 
just re-expressing an existing resource in TM syntax doesn't add any 
value.  One use case I can see is where you grab additional assertions 
about an existing topic from a RDF source (whether by query or by 
dereferencing a PSI/URL), and use those to enrich your topic map.

In general, I would think this approach is more realistic for delivering 
RDF subgraphs as TM fragments than for trying to convert a complete RDF 
resource like DBpedia to TM syntax.  IOW, it should be a dynamic 
process.  Bear in mind that the RDF itself may be generated from a 
further (dynamically changing) source, such as database content 
extracted by a Virtuoso Sponger Cartridge.

>> I am trying to understand what you think others could contribute to the
>> DBpedia topic maps project? Assuming it became a public project.
>
>It comes down to a simple RDF to Topic Maps mapping. We take the
>DBpedia ontology and decide how we'd translate it to Topic Maps.

Isn't that what Mappify already does? [1]

Richard

[1] 
http://ws.mappify.org/wp/url/http%3A%2F%2Fdbpedia.org%2Fresource%2FBremen

-- 
Richard Light