emergency-rim — archive

[Rex Brooks]

Hi Folks, In our last meeting I said that I thought an OWL 2 representation in the Manchester Syntax could be more efficient for data handling. Jeff asked what I meant, and I had no answer, due to an attack of brain freeze and the fact that the meeting was about to adjourn and my mind was on other subjects. So I thought I would clarify that statement because I happen to be studying OWL and the Manchester Syntax as I continue my work on the EDXLData-Domain Model in Protege. The elusive memory my brain wouldn't come up with automatically was that I had heard, most likely in the Ontolog Forum, or from another member of that group active in NCOIC, that the Manhcester Syntax works with the xsd 1.0 datatypes, so typing data as from that group of definitions would make the work of using the ontology in data-dependent applications (and implementations of standards such as ours which use the xsd datatypes) easier. Well, I don't really know whether that is true or not since I haven't tried it. And trying it was on my agenda, and bubbled up to the top today. Hence I'm reading through the OWL 2 Primer and the OWL 2 Specification, and other related materials. And what I have discovered further reinforces the notion of improved data handling in OWL 2. The more relevant reason for this is that the structural specification for OWL2 (one among several specifications focused on various aspects of OWL2) is defined using UML, which also lends itself to database specifications. Whether that is true or not, I still have to try it. One more factor is that OWL 2 is made to use the larger set of xsd-specified datatypes in XSD 1.1 which are optional for now until W3C finally adopts XML 1.1 as a Recommendation. However, whether it is possible or easy to include an OWL 2 specification of EDXL DATA-Domain Model as well as OWL1, which is the name given to the original specification for simplicity in reference. For now all I can say is that it is interesting, and I wish I had more time to devote to this. Unfortunately due to the press of other interests that need to have the EDXL Data-Domain Model to work with, I have to give priority to pushing forward on the class-based version. If there is an easy or automatic way to create an OWL 2 version, that would be great and I would argue in favor of developing both in tandem. Cheers, Rex