In what sense are DCMI Metadata Terms "based on RDF"?

[tombaker]

For over twenty years DCMI has sought to balance the need to clearly state that DCMI metadata terms are based on RDF with a need to point out that the standard is also, in some sense, technology neutral.

In the 2000s, it was controversial within DCMI to commit too strongly to RDF. The DCMI Abstract Model was developed between 2003 and 2007 as an RDF-compatible replacement for a metadata model that had emerged from DC workshops and conferences between 1996 and 2001 and that was unique to DCMI.

The Abstract Model never found wide adoption, perhaps because it was seen as too formal and RDF-like for people who were used to seeing metadata as ad-hoc document structures (eg, MARC, XML) while not being recognized by the Semantic Web community as something clearly based on RDF. Instead of bridging a gap, it fell between two stools.

Might we formulate a public position on this question to be featured prominently on the website?

[tombaker]

The DCMI Metadata Terms specification says:
explanation:

DCMI metadata terms are expressed in RDF vocabularies
for use in Linked Data. Creators of non-RDF metadata
can use the terms in contexts such as XML, JSON, UML,
or relational databases by disregarding both the
global identifier and the formal implications of the
RDF-specific aspects of term definitions. Such users
can take domain, range, subproperty, and subclass
relations as usage suggestions and focus on the
natural-language text of definitions, usage notes,
and examples.

The Introduction to ISO 15836 Part 2 says:

This set of properties and classes is expressed as an RDF vocabulary
and may be used in Linked Data. Each property and class has a global
identifier (URI).

The ISO 15836 glossary defines terminology used in the standard in that way that makes explicit their basis in RDF. "Class" is defined as follows:

The members of a class are known as instances of the class. Classes
are themselves resources. They are often identified by URIs and may
be described using RDF properties. The rdf:type property may be used
to state that a resource is an instance of a class.
[SOURCE: RDF Schema 1.1]

And "property" is defined as:

relation between subject resources and object resources

Note 1 to entry: This is a synonym of “element”.
[SOURCE: RDF Schema 1.1]

And "subproperty":

property that is related, typically using the rdfs:subPropertyOf
property, to another property typically of broader scope
(superproperty), such that all resources related by the subproperty
are also related by the superproperty
[Source: RDF Schema 1.1]

In addition, each property and class in the ISO standard has a URI, and any domains, ranges are specified with URIs.

In addition, ISO 15836 Part 2 has an Annex B (informative), "Dublin Core metadata as linked data", which explains:

Dublin Core metadata may contain three kinds of links:

1. URIs of the described resources themselves (such
   as works and their manifestations)

2. URIs of the schema elements

3. URIs of the values (such as names, subject headings)

Resource URIs may be based on standard identifiers
such as International Standard Book Number (ISBN) for
books or International Standard Text Code (ISTC) for
textual works. They may be used for providing
persistent links to the resources or metadata about
them. Identifiers are also an efficient means for
linking for instance work and manifestation metadata
records of the same resource, or for linking
interrelated works or manifestations.

Schema element URIs may be used to link the data
element (term) into its description, in order to
foster machine (and human) readability of the
metadata.

Value URIs may be used for creating links from
element values (such as Albert Einstein as an author)
to relevant Internet resources. URIs are either
Uniform Resource Locators (URLs) or persistent
identifiers such as Uniform Resource Names (URNs).
Persistent identifiers may be linked to 1-n Web
resources via a PID resolution process.

Dublin Core standard specifies only the links of the
2nd type. Linking to either resources themselves or
term values is beyond the scope of this standard.

[tombaker]

Tom and Niklas agree that it would be useful to have a statement we can point to from various pages and documents on dublincore.org, so we intend to polish this and post it to the Usage Board for discussion and approval.