more identifiers
Bruce D'Arcus
identifier variables.
<http://bitbucket.org/bdarcus/csl-schema/issue/17/identifiers>
.. but since this necessarily deals with data modeling issues, thought
I'd run it by here before moving forward.
Bruce
Bruce D'Arcus
For example, do we add a property for ...
<http://en.wikipedia.org/wiki/United_Nations_Document_Codes>
...?
Bruce
Frederick Giasson
>> So I have a ticket open for the CSL schema on adding some new
>> identifier variables.
>>
>> <http://bitbucket.org/bdarcus/csl-schema/issue/17/identifiers>
>>
>> .. but since this necessarily deals with data modeling issues, thought
>> I'd run it by here before moving forward.
>>
>
> For example, do we add a property for ...
>
> <http://en.wikipedia.org/wiki/United_Nations_Document_Codes>
>
Maybe we could add main identifiers in the core namespace, and create a
new namespace for an extensive list of other namespaces. Or we use the
same namespace, but split them in different documents/pages.
Thanks,
Fred
sieh...@googlemail.com
> Maybe we could add main identifiers in the core namespace, and create a
> new namespace for an extensive list of other namespaces. Or we use the
> same namespace, but split them in different documents/pages.
The are almost as many identifiers as people and organizations that
deal with bibliographic data so would should not pollute the namespace
with more sub-properties of bibo:identifier. The question whether an
identifier is 'main' or not depends on time, region and other context.
I would also favour to move all existing identifier properties to a
namespace. This namespace would be a "registry of identifier types
recognized by the BIBO community" and a subset of all sub-properties
of bibo:identifier that are used anywhere on the Semantic Web.
Collecting and defining identifier types of document and collections
has been and will be done by many different people and organizations,
I don't see this as core goal of BIBO. There is a dream of replacing
all identifiers with URIs but I don't see this happen in real life.
By the way what is the purpose of bibo:uri ? What is the difference
between
my:document bibo:uri $URI .
and
my:document owl:sameAs $URI .
or is owl:sameAs not the right property? Using a property to point to
an URI that is packet in a literal feels somehow against the spirit of
RDF.
Cheers,
Jakob
bibo:
every URIs, so putting pollute
Bruce D'Arcus
<sieh...@googlemail.com> wrote:
> Frederick Giasson wrote:
>
>> Maybe we could add main identifiers in the core namespace, and create a
>> new namespace for an extensive list of other namespaces. Or we use the
>> same namespace, but split them in different documents/pages.
>
> The are almost as many identifiers as people and organizations that
> deal with bibliographic data so would should not pollute the namespace
> with more sub-properties of bibo:identifier. The question whether an
> identifier is 'main' or not depends on time, region and other context.
> I would also favour to move all existing identifier properties to a
> namespace. This namespace would be a "registry of identifier types
> recognized by the BIBO community" and a subset of all sub-properties
> of bibo:identifier that are used anywhere on the Semantic Web.
>
> Collecting and defining identifier types of document and collections
> has been and will be done by many different people and organizations,
> I don't see this as core goal of BIBO. There is a dream of replacing
> all identifiers with URIs but I don't see this happen in real life.
So you're suggesting removing all existing identifier properties and
moving them to a new namespace? Is that practical?
Also, how are we socially going to organize such a "registry"?
> By the way what is the purpose of bibo:uri ?
This mainly comes out of current citation practice, and is just to
emphasize digital artifacts with a sort of canonical uri.
Bruce
Ross Singer
(http://info-uri.info/registry/docs/misc/faq.html) for these n+1
identifier schemes? Some mechanism would need to exist to "register"
the scheme (although that could become the responsibility of whoever
wants the identifier), but then there's a parseable format to
recognize what the identifier stands for (without extending a
vocabulary umpteen million times).
If an HTTP URI eventually comes along to replace the info-uri, so much
the better, but at least it would be something usable in the meantime.
-Ross.
> --
> You received this message because you are subscribed to the Google Groups "Bibliographic Ontology Specification Group" group.
> To post to this group, send email to bibliographic-ontolog...@googlegroups.com.
> To unsubscribe from this group, send email to bibliographic-ontology-spec...@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/bibliographic-ontology-specification-group?hl=en.
>
>
>
>
sieh...@googlemail.com
> > Collecting and defining identifier types of document and collections
> > has been and will be done by many different people and organizations,
> > I don't see this as core goal of BIBO. There is a dream of replacing
> > all identifiers with URIs but I don't see this happen in real life.
>
> So you're suggesting removing all existing identifier properties and
> moving them to a new namespace? Is that practical?
>
> Also, how are we socially going to organize such a "registry"?
I don't ask for more identifiers in Bibo but to seperate them from the
rest of the ontology. There could be a registry on the Bibo website
that regularly harvests a set of RDF sources and presents the result
of the following queries
SELECT ?s WHERE {
?s rdfs:subPropertyOf bibo:identifier .
}
Instead of documenting identifier schemes, bibo should only link to
them or make them link to bibo via
rdfs:subPropertyOf bibo:identifier.
Ross wrote:
> While I'm loathe to suggest this, what about using info-uris
> (http://info-uri.info/registry/docs/misc/faq.html) for these
> n+1 identifier schemes?
This registry includes the very important metadata field
(unfortunately not as RDF property yet) 'namespace-authority'. The
most important information about an identifier schema is who manages
the identifiers. For LCCN it's the Library of Congress, for ISBN it's
the International ISBN Agency, for URN it's IETF etc. How about a new
property:
bibo:identifierAuthority rdf:type owl:DatatypeProperty ;
rdfs:label "identifier authority"@en ;
rdfs:comment """The organization responsible for assigning
identifiers of a given identifier namespace and/or defining rules how
to assign identifiers in this namespace."""@en ;
rdfs:domain bibo:identifier ;
rdfs:range foaf:Organization .
Bibo should not define zillions of types of identifiers but define how
to document new identifier types for bibliographic material and give a
few examples (the identifiers that are already included in bibo).
Jakob
P.S: If you are happy with putting identifier types directly in bibo,
I have a couple of new identifier types that I need properties for.
sieh...@googlemail.com
> > By the way what is the purpose of bibo:uri ?
>
> This mainly comes out of current citation practice, and is just to
> emphasize digital artifacts with a sort of canonical uri.
But the URI is already there unless the document is a blank node! If I
understand it right, bibo:uri does not contain 'an URI of a document'
but 'a canonical URI of a document intended for citation'. That's a
difference which should be documented! Otherwise every known document
has at least one bibo:uri. We should stress that
{ $DOCUMENT bibo:uri $URI } => { $DOCUMENT owl:sameAs $URI }
But not the other direction.
Jakob