- From: Thomas Passin <tpassin@tompassin.net>
- Date: Thu, 29 Nov 2018 07:59:40 -0500
- To: thomas lörtsch <tl@rat.io>
- Cc: Pat Hayes <phayes@ihmc.us>, SW-forum Web <semantic-web@w3.org>
I agree. I'd add that these built-in features can be modeled in RDF (because ultimately they can be diagrammed as boxes connected by lines). My suggestion is that for the b-node cases in point, it might be a good idea to model the notion of a Topic Map "scope". IOW, use the TM feature as an RDF idiom. TomP On 11/29/2018 3:56 AM, thomas lörtsch wrote: > I guess the gist of it is that Topic Maps are higher level than RDF, they have more semantics build in. You define classes with certain properties like preferred names and alternative names and associaztions between classes and scopes on certain attributes. Then you add occurrences of those classes. The model is nowhere as flexible as RDF which is both its strength and its weakness. It is tailored to specific use cases - like e.g. aligning documentations from different sources - and provides a very nice primitive/template/idiom for them. There was some work on aligning Topic Maps and RDF and one of the most mature was I think Lars-Marius Garshol’s 'Q-model' - Q like in quints [0], but he tried quads too. > > In those days I liked Topic Maps much more than RDF precisely because they had that notion of scope. They also got the distinction between subject indication and denotation right from the start. That was one of the main problems when trying to align them with RDF. > > Thomas (L.) > > > [0]http://www.ontopia.net/topicmaps/materials/the-q-model.pdf
Received on Thursday, 29 November 2018 13:00:08 UTC