Re: [ontoiop-forum] Using ISO 11179 to specify file format of a remote o

["terry.longstreth" <terry.longstreth@xxxxxxxxxxx>]

Christoph,
Thanks for the question.  Sorry it's taken me this long to get back to you, but my writing process is proto-Proustian.  I find it hard to extemporize in writing, because any text preserves my ignorance for eternity. 

 To start, I'd recommend reviewing the recent 2012 Metadata OpenForum  (http://www.metadata-standards.org/OpenForum2012/presentations.html).

My concern was that it's difficult for readers of standards documents to disassociate neutral examples from recommendations for specific strategies.  By mentioning the linked data meme and HTTP header, you we're (intentionally or not) suggesting that that was one good way to capture metadata for any (linearly expressed object expressing an) ontology.

I don't think the case has been made that such a strategy will be effective in the long run. 

I do know that there are infinitely other ways to associate metadata with objects. I'm personally familiar with metadata packaged in an ASN.1 Management Information Base (MIB), an HTTP header, or (my preference) a standard reference location such as those that could be provided by an ISO 11179 conforming Metadata Repository (MDR) or  MDR family as described in ISO 20944- Metadata Registry interoperability & binding.  I concede 11179 is rather abstract, but I think that's appropriate for our case.  For example, the ISO 19763, Metamodel Framework for interoperability builds upon the ISO 11179, and supplies some sharable (it is hoped) metadata schemas for use in a specific context.

I emphasize that I neither know nor care particularly what method or strategy is chosen.  Another standards activity that I'm involved in is that of ISO TC20/SC13's Open Archival Information Systems (OAIS).   The OAIS Reference Model  (ISO 14721 AKA  CCSDS Recommendation 650.1) is considerably more abstract than the ISO 11179, but it supplies common terminology and a shared conceptual framework to be used by digital archives as concerns long-term preservation and usability of the archived data.  We're adding standards based on OAIS  for Audit and Certification of such archives.  For more on this effort I invite you to browse our open access Wiki.

I hope this has helped to clarify my concerns. I know it sounded / looked like a Nit on the chat.
Regards,
Terry


Christoph LANGE wrote:

Hi Terry,

let me use the public mailing list for a question that follows up on our discussion in the chat during today's teleconference.

We were discussing the issue of how to inform an application that one ontology O within a distributed ontology D is serialized in a certain format.

When the ontology O is written in place inside D we have the keyword "serialization" for this.

When O is a remote ontology referenced by IRI, we agreed on something I'd like to phrase as follows:

1. If information on the serialization of O is accessible in some standard way (but standardizing this is beyond DOL), an application should make use of this information.  E.g., and this is something that could be mentioned in an informative note, O could be published according to the linked data best practices, such that it is downloadable from its IRI=URL, and the Content-type HTTP header could contain information that the ontology has the text/owl-manchester MIME type.

2. If no such information is provided, it is up to the application how to determine the serialization of O.  E.g., if the ontology is downloadable from its IRI=URL but the HTTP headers don't hint at the serialization, the application could download the file and apply some heuristics to it in an entirely application-specific way.

OK, and now concerning (1) you made me aware that my example is not the only way of providing the "serialization" information, but that ISO 11179 could be used as well.  (BTW, some non-standard RDF property could also be used, in an RDF triple like <IRI-of-O> <http://some.ontology/isAvailableInMIMEType> "text/owl-manchester".)

ISO 11179 "Metadata registries" part 1 (framework) and 2 (classification) all look quite abstract to me.

Could you explain to me how (1) could be solved using ISO 11179?

Cheers, and thanks in advance,

Christoph

--

_________________________________________________________________
To Post: mailto:ontoiop-forum@xxxxxxxxxxxxxxxx
Message Archives: http://ontolog.cim3.net/forum/ontoiop-forum/  
Config/Unsubscribe: http://ontolog.cim3.net/mailman/listinfo/ontoiop-forum/ 
Community Files (open): http://interop.cim3.net/file/pub/OntoIOp/
Community Wiki: http://ontolog.cim3.net/cgi-bin/wiki.pl?OntoIOp    (01)