W3C home > Mailing lists > Public > public-rdf-shapes@w3.org > July 2014

Re: Shapes/ShEx or the worrying issue of yet another syntax and lack of validated vision.

From: Dimitris Kontokostas <kontokostas@informatik.uni-leipzig.de>
Date: Fri, 18 Jul 2014 07:41:31 +0300
Message-ID: <CA+u4+a0B8Mv9FvcVxSMfmUU_9TCFuypRr+K4PqwS0goZ6GLjkA@mail.gmail.com>
To: Karen Coyle <kcoyle@kcoyle.net>
Cc: Simon Spero <sesuncedu@gmail.com>, "public-rdf-shapes@w3.org" <public-rdf-shapes@w3.org>
On Fri, Jul 18, 2014 at 6:25 AM, Karen Coyle <kcoyle@kcoyle.net> wrote:

> On 7/17/14, 3:58 PM, Simon Spero wrote:
>> On Thu, Jul 17, 2014 at 6:26 PM, Karen Coyle <kcoyle@kcoyle.net
>> <mailto:kcoyle@kcoyle.net>> wrote:
>>     Absolutely. In fact, I doubt if anyone can guarantee that their data
>>     will never have to interact with OW assumptions. I do hope that we
>>     keep in mind that we must support both contexts.
>> Karen -
>>     to help clarify discussion, could you give some possibly give some
>> specific examples from the Library, Archive, or Museum domains where the
>> OWA is especially problematic?
> I actually think I was saying the opposite: that most LAM data sources
> that I encounter are intended for public consumption, and therefore try to
> be aware of the OW and what it means for their data. At the same time, data
> creators, and those LAM institutions that are exchanging data amongst them
> for their private purposes, have a need to create data that validates
> against certain rules. Two examples are the data aggregators Europeana and
> Digital Public Library of American. They aggregate data created in widely
> diverse contexts and that do not follow a common data format, and therefore
> want to validate incoming data. The rules used for that validation (one and
> only one title, as an example) are not ones that would apply to the OW.
> As for an area where OWA is problematic, one need look no further than
> FRBRer ontology[1], which is clearly designed using OWL constraints (which
> I prefer to call "axioms" to avoid confusion) in a closed world way. The
> definitions are quite strict, with all classes disjoint each other, such
> that, using reasoning in the OW, any FRBRer data will be inconsistent with
> data not using that exact set of axioms.

We used a FRBRer dataset as a use case in the main RDFUnit paper [1] and we
actually came to the same conclusions.
As mentioned earlier, we prefer SPARQL property paths for type inferencing
since reasoning can hide or introduce new inconsistencies.


[1] http://svn.aksw.org/papers/2014/WWW_Databugger/public.pdf

> Did that answer your question? Or did I misunderstand your purpose?
> kc
> [1] http://iflastandards.info/ns/fr/frbr/frbrer/
>> Thanks.
>> Simon
> --
> Karen Coyle
> kcoyle@kcoyle.net http://kcoyle.net
> m: 1-510-435-8234
> skype: kcoylenet

Dimitris Kontokostas
Department of Computer Science, University of Leipzig
Research Group: http://aksw.org
Received on Friday, 18 July 2014 04:42:28 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 17:02:39 UTC