Re: meeting

[James R. Griffin III]

Dear Eric,

1) It seems a general fedora/hydra concern is changing predicates.  We are having the issue with the bounding box.  It seems a useful function of the repository would be a simple (in concept not implementation) of a repository wide replace ? Or, as I am learning from the SemWeb book, using OWL equivalencies?

Thank you for bringing this topic to the forefront.  My apologies, but I was unaware that there was a predicate being changed.

We are having the issue with the bounding box.  It seems a useful function of the repository would be a simple (in concept not implementation) of a repository wide replace ?

While in principle I favor this, I believe that there would be not suitable replacement (we would essentially need to supply our own predicate, resolving to a geo_concerns-specific namespace, or, opaquenamespace.org).

Or, as I am learning from the SemWeb book, using OWL equivalencies?

To the best of my understanding, this equivalency still requires a valid URI:

<owl:Class rdf:ID="FootballTeam">
  <owl:sameAs rdf:resource="http://sports.org/US#SoccerTeam"/>
</owl:Class>

Because no such valid URI exists, such a relationship could be encoded/structured, but never properly resolved.  Please correct me if this is incorrect, or if I am misunderstanding this question (I am working from https://www.w3.org/TR/owl-ref/#equivalentClass-def and https://www.w3.org/TR/owl-ref/#equivalentProperty-def).

Best,

James

On Tue, Jan 19, 2016 at 2:01 PM, James, Eric <eric....@yale.edu> wrote:

Hi group,

Sorry I am away from the office, was lurking in the hangout with no mic.  Been away from the geo_concerns front for a while, I apologize I just have a couple general questions.

1) It seems a general fedora/hydra concern is changing predicates.  We are having the issue with the bounding box.  It seems a useful function of the repository would be a simple (in concept not implementation) of a repository wide replace ? Or, as I am learning from the SemWeb book, using OWL equivalencies?

2) Another wish list thing, not sure if it does this as is: adding functionality to the geo concerns so uploading metadata would autocomplete the metadata form fields where applicable.

I will be and geo4lib4 camp next week too, looking forward to fleshing things out more there.

-Eric

--

James R. Griffin III
Digital Library Developer
Digital Scholarship Services
111C Technical Services
David B. Skillman Library
Lafayette College
Easton, PA 18042

[James, Eric]

James,

Thanks for your response.

The case, which I think is similar to the one with the geoRSS box is, say you publish a URI, perhaps not a vetted one, and find a more suitable one later.  My understanding is that one could 1) replace from repository/triplestore, or 2) use OWL to express that the replacement is equivalent to the original by reciprocal equivalence.  Again correct me if I'm wrong, one could use any made up URI of valid format in an ontology, (a URI doesn't have to resolve to anything to be used, it's using it that gives it it's value).

-Eric

---

From: hydra-gis-w...@googlegroups.com [hydra-gis-w...@googlegroups.com] on behalf of James R. Griffin III [jrgrif...@gmail.com]
Sent: Tuesday, January 19, 2016 2:41 PM
To: James, Eric
Cc: Darren Hardy PhD; Eliot Jordan; John Huck; hydra-gis-working-group
Subject: Re: meeting

--
You received this message because you are subscribed to the Google Groups "Hydra GIS Working Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hydra-gis-working...@googlegroups.com.
To post to this group, send email to hydra-gis-w...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/hydra-gis-working-group/CAHJkeYAq3WDxjMG1V7cjf2WE1ci1OGkuQrM84ayd2yuUCFRSkw%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.

[James R. Griffin III]

Dear Eric,

 My understanding is that one could 1) replace from repository/triplestore, or 2) use OWL to express that the replacement is equivalent to the original by reciprocal equivalence.  Again correct me if I'm wrong, one could use any made up URI of valid format in an ontology,

Absolutely, this is my understanding as well.

say you publish a URI, perhaps not a vetted one

This was precisely where my previous response was flawed, sorry for that.  My approach, thus far, has been to reference only fully-vetted published URI's.  I have been attempting to contact the Hydra Metadata Interest Group (both the Applied Linked Data WG, and the IG itself) in order to determine the best approach to publishing such a URI using a schema (RDFS or OWL) for an ontology which we've developed.

and find a more suitable one later

 

Agreed, I'm certainly in favor of this.  Also, I believe that this would ensure that this would suit the the approach outlined by Eliot earlier regarding geoRSS box.

Returning to your initial question:

[...] a simple (in concept not implementation) of a repository wide replace ? Or, as I am learning from the SemWeb book, using OWL equivalencies?

I would please be in favor of working to implement a solution which uses OWL equivalencies.  RDF::Vocab seems to offer some methods through which to dynamically declare such relationships:

https://github.com/projecthydra-labs/rdf-vocab/blob/master/lib/rdf-vocab/vocab/mads.rb#L159

I'm uncertain as to whether or not it would be the best practice to attempt to integrate such functionality within a Module for GeoConcerns, but perhaps the design patterns which have been used here could still be of use.

Sincerely,

James

[John Huck]

I thought I saw you on the hangout, Eric, but I couldn't be sure. Nice to know you are working through the Sem web book as well.

I think you and James have arrived at the same page. At the risk of repeating what you've both already said, the issue, as I understand it, is that the bounding box predicate was never a valid term to begin with, because the URI is not dereference-able and so it's, I guess you could call it a namespaced string (? I just made that up) rather than an RDF resource. In an OWL document, I think you're right, it would matter less that they weren't dereferenceable. There would be some way to make explicit the relationship between the RDF resources (i.e., the new predicates) and the non-RDF entities (the GeoRSS elements) that they were based on.

John

--

John Huck

Metadata & Cataloguing Librarian

University of Alberta Libraries

5-25E Cameron Library

(780) 248-1337