- From: Jacek Kopecky <jacek.kopecky@deri.org>
- Date: Fri, 28 Sep 2007 15:00:49 +0200
- To: Richard Cyganiak <richard@cyganiak.de>
- Cc: mhepp@computer.org, "Semantic-Web@W3.Org," <semantic-web@w3.org>, Katharina Siorpaes <katharina.siorpaes@deri.at>
On Fri, 2007-09-28 at 13:19 +0200, Richard Cyganiak wrote: > If you really want to make sure that all agents encountering your > data work off the same vocabulary definitions, then you should > probably duplicate the relevant terms in your own namespace, creating > hepp:name and hepp:knows and so on, and declare them > owl:equivalentProperty to the original terms. FWIW, the way I interpret Martin's situation (and I'll be glad for corrections) is that he wants some specific tools to only see the portion of the ontology, not all possible tools in the wild. For instance, a tool that generates forms from an ontology for easy input of instances would only see a few DC or FOAF properties and the resulting forms would be easier and more appropriate to the given application. So I'd suggest just doing the copy&paste (with relevant disclaimers for copyright, I think this is fair, though), testing that it works in the tool in question, and being done with it. As long as dependencies are taken into account in the copy&paste (as discussed by Alan), the resulting data should also interoperate well with the rest of SW data, which for me would be the biggest concern. Therefore I'd say it's OK for Martin to subset DC and FOAF, it's not like he's claiming to have the complete and authoritative definition of DC and FOAF in his subsets, or that there would be risk of tools being fooled into taking the subsets as the complete and authoritative defns. Best regards, Jacek
Received on Friday, 28 September 2007 13:01:20 UTC