Re: XML Europe 2003 Call for Participation

On Thursday 24 October 2002 21:58, Danny Ayers wrote:
> >>While topic map technology has always had an excellent showing at this
> >>conference, representation from the RDF world has always been thin.
> >>I'd love to see more proposals from the RDF world for XML Europe 2003,
> >>so here's your chance.
> >
> >Fully agreed, flavored by a little curiosity. Whereas RDF represents a
> >fairly streamlined datamodel, constituting representational elegance, and
> >a strong push toward interoperable ontologies, I would see topic maps as
> >having a complex set of concepts aimed at little more than what vanilla
> >XML can do. That is, topic maps do not incorporate strong semantics,
> >whereas the RDF community embraces a strong push towards making its
> >semantics unambiguous. To me this suggests that topic maps are little more
> >than an extra transfer syntax, while things built on RDF (another transfer
> >syntax/data model) hold a much broader promise.
>
> I've only recently starting having a proper look at TMs, but from what I've
> seen so far I have to disagree - most of the plusses you give RDF do also
> apply to TMs. Their area of application is a little to the side of RDF, and
> in terms of versatility I think it's reasonable to say RDF has the upper
> hand.

I think that even I as a hardened topic map addict would have to agree with 
you on this point. Of course, I'm having difficulty thinking of a concrete 
example ;-). Seriously, I think that topic maps have little to offer those 
looking at an RDF vocabulary such as DAML+OIL and so, for a certain set of 
applications, you would be forced to 
a) create a new vocabulary from scratch
b) twist the topic map model to fit a description logic view of the world

> But I really don't think 'vanilla XML' is a fair description at all -
> for example their approach to 'URI as concept' vs 'URI as page' actually
> seems a lot better defined than that of RDF. 

Yes, and rather than sneer at topic maps, I would suggest that the RDF world 
might consider looking at some of the core topic map principles and 
determining what might make a useful subset for application as a simple RDF 
application. This is a line of thought which I have been persuing for a 
while, and hope to have written up as a paper in the near future.

> I would suggest that TMs
> probably have a great deal to offer the Semantic Web, and interoperation
> with RDF isn't a difficult problem because of the commonality of URIs
> (amongst other things).
>

Its almost trivial IMO.

> >Against this background, it's peculiar at the very least that topic maps
> >would be the technology to prevail in XML Europe. In XML Finland, it
> >seemed that Semantic Web rather took the show. If TM really has been the
> >king of XML-E, it's high time the tide turned.
>
> A (somewhat cynical) explanation would be that although TMs offer less than
> RDF, they've already delivered. 

It is also a curious instance of the commercial application developers getting 
the drop on the open-source community. They have brought a variety of topic 
map solutions "to market" in a remarkably short space of time. I am not 
suggesting in any way that this is a factor in there representation in the 
conference, but companies producing topic map toolkits and applications are 
very well represented on the show floor at the GCA conferences. 

> Whatever, I don't think anyone has to make
> an exclusive choice between RDF and TMs, and diversity where there is
> interoperability can only be a good thing for the SW. So the turn of the
> tide is only really an issue for King Cnut.
>
Exactly. No one ever said that the SW had to be built out of one technology. 
In fact, just imagine how boring it would be if it were.

> Having said all that - I'm right with Edd, let's see some more RDF at these
> places ;-)

Yeah, both in the conference rooms and on the show floor please!

Cheers,

Kal

-- 
Kal Ahmed, techquila.com
XML and Topic Map Consultancy

e: kal@techquila.com
p: +44 7968 529531
w: www.techquila.com

Received on Tuesday, 29 October 2002 16:08:30 UTC