Thanks Jeff for setting me straight on Schema/OWL sameAs, and how reasoners handle them differently. 

Somewhere along the way I picked up the unfounded idea that schema:sameAs had an equivalent object property assertion with owl:sameAs. Not the case. 

Forgive any typos, sent while on the run.

On Sep 8, 2015, at 11:38 PM, Jeff Young <[log in to unmask]> wrote:

Steven,

I think I can prove that the semantics I describe are consistent with RDFS, OWL, Schema.org and SKOS. Neither RDFS nor OWL reasoners are going to make inappropriate inferences by confusing schema:sameAs with owl:sameAs. The pattern I describe is even consistent with these SKOS constraints:

http://www.w3.org/TR/skos-reference/#S14
http://www.w3.org/TR/skos-reference/#L4858

I agree that statements like the following are wrong, but are probably inevitable until the notion of "authority" fades away. 

<Some person> madsrdf:editorialNote "Non-Latin script references not evaluated.” .

If LC authorities replaced their blank nodes with URIs or reconceptualize their current URIs to be RWOs, VIAF could help push some of the authority weirdness further into the background.

Jeff

On Sep 8, 2015, at 11:01 PM, Steven Folsom <[log in to unmask]> wrote:

Jeff,

Thanks for the response (and all that OCLC has done to progress the linked data agenda in the library community). I didn’t mean to call anyone out.

I agree the RWO/authority conversation is difficult, perhaps the hardest thing to wrap heads around when we’re so used to talking/modeling in terms of authorities… especially now that RDA allows us to say more explicitly about the RWO. I’m happy that OCLC uses separate URIs for the different entity types, but it seems like a big concession not to maintain consistent semantics. [An aside, everyone should read Jean Godby, Shenghui Wang, and Jeff Mixter’s book on linked data. It has the most accessible description of how authorities and RWOs differ.]

One of the motivations for migrating to linked data/RDF is to benefit from the formal logic that RDFS and OWL provide; we want to benefit from inferred data. The sameAs assertion between a person and authority means if we run reasoning over our data we’re going to end up with triples like:

<Some person> madsrdf:editorialNote "Non-Latin script references not evaluated.” .

(I’m sure someone can come up with more awkward examples; that’s not a challenge. :)

LOC has blank nodes for RWOs see: http://id.loc.gov/authorities/names/n94112934.rdf. They are even typed appropriately with FOAF persons, organizations, etc. now!!! It would be great if VIAF authorities could have reciprocating relationships with LC authorities, and VIAF RWOs could have reciprocating relationships with LOC RWOs.



Thanks,
Steven

————
Steven Folsom
Metadata Strategist and Standards Advocate
Cornell University Library







On 9/8/15, 2:06 PM, "Bibliographic Framework Transition Initiative Forum on behalf of Young,Jeff (OR)" <[log in to unmask] on behalf of [log in to unmask]> wrote:

Picking up a thread from a couple of weeks ago...

On Aug 26, 2015, at 11:44 AM, Steven Folsom wrote:
I noticed throughout this thread
(and in the VIAF data) authority URIs being used as the objects of
properties where instead person, places, etc. URIs should be. For
example, (as previously
mentioned) VIAF has separate URIs for the person and the authority,
but then they go on to make a sameAs assertion between the VIAF
*person* and an id.loc.gov *authority*. I hope this is just a mistake,
and that they would be open to making the sameAs assertion between the
VIAF foaf:Document and the id.loc.gov authority.

Here's a rationalization of VIAF's RDF treatment of *real-world entity* vs. *authority* ("sameAs" vs. "focus" vs. "identifiedBy", etc.)

- The VIAF URI (e.g. http://viaf.org/viaf/36997809) identifies the *real-world entity*
   - To maximize cross-domain interoperability, VIAF uses Schema.org vocabulary as much as possible to describe the entity.
       - E.g. http://bit.ly/1NKEvOa
       - These VIAF entities are not dependent on preconceptions of "authority".
   - VIAF entities use schema:sameAs rather than owl:sameAs/madsrdf:isIdentifiedBy to avoid splitting the entity/authority hair when linking to external URIs.
       - The use of schema:sameAs "tends to blur distinctions which are important to OWL users, such as URIs for entities versus the pages that describe them". In VIAF's case, the blurring effect is between *entity* vs. *authority*. This blurring is intentional because most consumers of VIAF entities won't understand or care about the difference.
           - http://www.w3.org/wiki/WebSchemas/sameAs
       - For use cases that *do* understand and care about the entity/authority difference, VIAF uses foaf:focus instead of schema:sameAs in the reverse direction.
           - E.g. http://bit.ly/1LInbWN
           - Unlike the VIAF *entities*, which use Schema.org vocabulary to maximize cross-domain interoperability, these *authorities* are mainly described using SKOS.

Hopefully this outline clarifies some of the details behind VIAF's RDF. If not, I can try to elaborate.

Jeff