-
Notifications
You must be signed in to change notification settings - Fork 18
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Should gist
describe its preferred prefix and namespace URI using VANN?
#684
Comments
I've used I'm putting this on the roster for consideration for the next release, as it's an easy and non-semantic fix. |
This makes a lot of sense to me, especially this:
|
Why do we have |
Reasons not to use:
Jamie: They facilitate machine readability. In conjunction they tell the machine what they should be so users don't have to. Rebecca: It doesn't matter what prefix people use. Michael: It would be perverse to use a prefix that is not the standard defined by the authors. Rebecca: The VANN ontology isn't retrievable from its ontology IRI (http://purl.org/vocab/vann/), either in Chrome or Protege. So clearly this isn't a commonly-used ontology and most machines would not be equipped to interpret these properties. Jamie: Not sold on VANN ontology for reasons above, but could use others. Perhaps SHACL - using in DCA. Class: |
@Jamie-SA Where did we leave this? Were you going to do any additional exploration, or is it still under group review? |
Here are the SHACL term definitions for everyone to review:
Here is their example of how to use them.
|
Michael/Rebecca: The SHACL terms aren't precisely what we want to express - they're used to associate a prefix with a namespace, not to specify the namespace itself. Could mint our own annotation properties. Dan: Could use the SHACL mechanism to specify multiple prefixes; e.g., the tbox, abox, and cbox patterns. Jamie: This way applications like visualizations could use the prefix declarations to match an IRI and display the prefix rather than the full namespace. Dan: Could put the SHACL prefix declarations into a separate file. This provides the guidance for applications without adding application-level cues to the ontology. Can add a bare PrefixDeclaration object in a separate file, without linking to the ontology With a blank node:
With an IRI node:
On gist:
Options:
What about the output of tarql? - Can add prefix declaration blank nodes as above if desired. Which prefixes would you declare?
DECISION:
If in future we decide to adopt the namespace pattern gist/gistx/gistd, we would change
|
Just catching up on this issue now(!), and I totally agree with the decision you folks arrived on here! But shouldn't the object of the |
This was completed and released. |
It's become very common practice in more recent years for ontologies to explicitly provide simple metadata for their preferred prefix and namespace URI using the long-standing, and stable, VANN vocabulary.
For gist, this would simply be including the following two new VANN triples:
Note: In the case of the prefix, this is just potentially helpful extra metadata - i.e., by explicitly stating a preferred value it can help maintain a higher degree of consistency among people using that ontology.
Note: In the case of the namespace URI, this can be helpful in linking the ontology resource (i.e.,
<https://ontologies.semanticarts.com/o/gistCore>
in the case of gist) back to the actual namespace URI it describes (i.e.,"https://ontologies.semanticarts.com/gist/"
).Note: The
rdfs:range
of the VANNpreferredNamespaceUri
predicate (although not explicitly stated in the VANN documentation) is accepted to be anxsd:string
, and not an IRI, orxsd:anyURI
.I think this is deliberate to allow processors perform simple string manipulation on the IRIs of terms from a vocabulary, but also because to-date the notion of a Namespace has not been explicitly denoted (i.e., there are no
rdf:Namespace
orowl:Namespace
types, and no associatedrdf:namespace
orowl:hasNamespace
predicates), and so dereferencing a Namespace URI in-and-of-itself has generally (but not always!) been expected to return the description of that ontology - but that description has never been expected to contain triples describing the Namespace itself (whether it should or not is a separate issue :) !).The text was updated successfully, but these errors were encountered: