Re: [ssn] equivalence axioms to relate the old ssn to the new ssn.

Hi,

Thanks for the comment and the clarification by Kerry.

The SSN sub-group had some robust discussions around the namespacing and modularisation of the SSN. The idea of the modularisation is to offer users different levels of ontological commitment, depending on their use case. From evidence gathered, few if anyone is using the strongest axiomatization in the current SSN, that is the DULCE layering in SSN. Therefore, the new SSN does not rely on this layering underneath DULCE, but there is a separate alignment file that ensures that users who rely on and prefer the strong axiomatization offered by DULCE are still able to use it. However, we decided to keep the alignment files separate in their own namespace, so that you do not get unintended inferences in your SSN namespace if the alignment ontology somehow ends up in your triplestore/reasoner. Also, on a Web scale, this will maintain a clear separation between the different layers of the SSN, including the new lightweight core (currently called SOSA), which is also in its own namespace with a different, more lightweight ontological commitment.

Cheers,
Armin

From: Kerry Taylor <kerry.taylor@anu.edu.au>
Date: Friday, 9 December 2016 at 10:41 am
To: Ghislain Atemezing-Pro <ghislain.atemezing@mondeca.com>
Cc: "public-sdw-wg@w3.org" <public-sdw-wg@w3.org>
Subject: RE: [ssn] equivalence axioms to relate the old ssn to the new ssn.
Resent-From: <public-sdw-wg@w3.org>
Resent-Date: Friday, 9 December 2016 at 10:42 am

Hi Ghislain,
Thks for your comment. I am very happy to see that someone is looking!

The idea to keep them in an ontology different to the main ssn ontology so that you don’t have to get them if you don’t want them (part of the modularity goal).  It seems that if you load an ontology into  lots of triple stores they load all the other ontologies referenced by term namespaces --- in this case that would mean loading the old (purl) ssn and the old dolce too (which is not even retrievable from that namespace any more, so really causes problems for tools). That’s pretty ugly behaviour.

So it’s a convenience thing. I don’t think the “namespace” is important --- it actually has no terms in the namespace (so I am not even sure that “namespace” is the right word here) but indeed the ontology file and the ontology URI would be different.

As a secondary reason --- making the equivalences a little less “in your face” is hopefully a way of encouraging the move to the new ssn.

Personally, I don’t feel much commitment to this way of doing it --- and there may well be a better way that I can’t think of. Or just putting it in the main ssn file could be the better one if this way creates problems for some people or tools.


And yes: http://www.w3.org/ns/ssn/  is considerably out of date.  Please see this instead https://github.com/w3c/sdw/blob/gh-pages/ssn/ssn_separated/ssn.owl . The version on the namespace location will not be updated until after the f2f vote (but before the formal publication of the WD).


Can you pls clarify “those nice metadata that you have in the equivalences file. ( + dct:issued property as well)” if the github version also suffers in that way?   And I will do what I can to fix it. I don’t recall dct:issues having been used at all before.


-Kerry




From: Ghislain Atemezing-Pro [mailto:ghislain.atemezing@mondeca.com]
Sent: Thursday, 8 December 2016 1:36 AM
To: Kerry Taylor <kerry.taylor@anu.edu.au>; public-sdw-wg@w3.org
Subject: Re: [ssn] equivalence axioms to relate the old ssn to the new ssn.

Hi Kerry,
Nice work! Those mappings are really necessary for end users.
However, I don't understand why we need a new namespace/ontology for this type of mappings.
Wouldn't be better to have those mappings in the same ontology file, in the new ssn namespace?
Btw, I was looking at http://www.w3.org/ns/ssn/ and there are still missing those nice metadata that you have in the equivalences file. ( + dct:issued property as well)

Best,
Ghislain

Le mar. 6 déc. 2016 à 14:52, Kerry Taylor <kerry.taylor@anu.edu.au<mailto:kerry.taylor@anu.edu.au>> a écrit :
SSN-ers,
I have built a small ontology to relate the classes and properties of ssn in its new namespace  (http://www.w3.org/ns/ssn/) to the classes and properties of the old  (SSN-XG) ssn in its old namespace .  The intention of this is that it becomes a part of the normative ssn but is served in a different ontology file so that it does not need to be loaded up in order to use the new ssn. However, it (hopefully) will enable us to use implementations of the old ssn as proof of implementations of the new ssn (where concepts by all of intension ,  rdfs: comment description,  and the suffix part of their names have not changed).

Please see it here https://github.com/w3c/sdw/blob/gh-pages/ssn/ssn_separated/ssn_equivalences.owl  I would like to include this in the FPWD so that our plan for implementation evidence can be tested before it is too late. I will also get it into the rec document, I hope.

Any comments gratefully received.
-Kerry

--
--------------------------------------------
Ghislain A. Atemezing, Ph.D
R&D Engineer
@ Mondeca, Paris, France
Labs: http://labs.mondeca.com

Tel: +33 (0)1 4111 3034
Web: www.mondeca.com<http://www.mondeca.com>
Twitter: @gatemezing
About Me: http://atemezing.org