[topicmapmail] Web Services

[Christoph Froehlich]

Am Do, den 15.01.2004 schrieb Thomas B. Passin um 02:11:
> Thomas Schwotzer wrote:
> 
> > Other application classes don't need such a fine grained access
> > and don't need transactions. Some applications might wish to access
> > a Topic Map like a Web Site. They ask for some information and get
> > a Topic Map fragment in return. SNAPI shall become such an API.
> > It is supposed to be a kind of HTTP e.g. for Topic Maps. There are
> > no transactions and also no way to access a particular topic.
> > XTM fragments can be exchanged. Advantage: Light-weight implementations.
> > 
> 
> There are two kinds of things to deal with in this arena, I think -
> 
> 1 - How should you ask for some relevant part of a topic map (so as to 
> avoid getting the whole thing when it is not needed)?
> 
> 2 - What is a "relevant" part of a topic map, anyway?  How do you know 
> where to stop?
> 
> I think the most promising approach for 2) - really the only one I know 
> about - is the "Concise Bounded Description" Proposed (for RDF) by 
> Patrick Stickler - see
> 
>     http://sw.nokia.com/uriqa/URIQA.html
> 

Another approach is taken by the panckoucke library (a subproject of
tm4j). When a distinct topicmap element is requested, panckoucke
generates and returns a model that represents the requested element.
What "represents" actually means, depends on a piece of plugable sotware
called the Abstractor. An Abstractor may implement the CBD, or a
statistical analysis or soemthing else. 

Panckoucke is currently used to feed presentation engines. Since some
time, I wonder, if it shouldn't be possible to define the model in a way
that roundtripping could be done. This would allow us to build editors
based upon the panckoucke abstract model.

For further reading, the current, spare docs are available online on
http://www.tm4j.org/panckoucke/docs/userdoc/

bye
christoph

 
> This is actually easier to do in topic maps than in RDF because we don't 
> have to do extra work to find and chase down all the anonymous nodes.
> 
> Basically, the CBD gives you all topics that are directly related to a 
> given topic.  Thus the results have a depth of just one.  To follow up 
> on any of them you ask for its CBD.  In RDF, it would be all triples in 
> which the given node is a subject, and all triples for all bnodes 
> connected to that node (continuing, if I have it right, until you have 
> exhausted each chain of bnodes),
> 
> I tried this out experimentally with my topic map engine.  It was pretty 
> easy and seems to work well.
> 
> Cheers,
> 
> Tom P
> 
> _______________________________________________
> topicmapmail mailing list
> topicmapmail@infoloom.com
> http://www.infoloom.com/mailman/listinfo/topicmapmail
-- 
Christoph Froehlich <cfauto@folge2.de>