Hi James,
Jumping back to your message of a few weeks ago...
> Over the weekend i began to read up more about the Dublin
> Core, its pretty new to me, although i have herd of it ive
> never actually implemented it into any of my Web Documents.
>
> After reading some of the documents at the Dublin Core site i
> got lost rather quickly. I understand that the DC Elements
> are implemented within the <meta> tags in HTML documents, and
> have strong knowledge of HTML, so on that side of things im
> pretty clear, but below i have a few questions. If someone
> could dedicate a few moments of their time in order to help
> me, i would be greatly thankful.
>
> 1 - When embedding DC elements in HTML documents, do i need
> to include any document or vocabulary before defining DC.
> elements? For example, HTML pages include a doctype to point
> to a set of rules, does the DC need anything like this within
> my documents?
If I understand your question correctly, I think what you are asking for
here is handled by the value of the X/HTML profile attribute i.e. the
inclusion of the profile attribute with a value of
http://dublincore.org/documents/dcq-html/
<head profile="http://dublincore.org/documents/dcq-html/">
</head>
is a signal provided by the X/HTML document creator to indicate to a
consumer/reader of the X/HTML document that the meta/link tags follow
these conventions and encode DC metadata.
> 2 - A DC element is NOT the same as a HTML element right?
> Example, a HTML element would be <meta> whereas a DC element
> is not <meta> but instead DC.title (the name trailing the
> period, 1 of 15 i believe).
To understand what is happening here, I think it's necessary to take a
step back. The nature of "Dublin Core metadata" is defined by the
specification called the DCMI Abstract Model:
http://dublincore.org/documents/abstract-model/
Amongst other things, that document defines what it calls a "description
set model" i.e. it defines an information structure which it calls a
"description set" and describes its components and their relationships.
The DCAM defines an "abstract" information structure, if you like; it
doesn't say anything about how an instance of that information structure
is to be represented in concrete form. That is the job of the
specifications which DCMI calls "encoding guidelines": they describe how
to "encode" a description set in some digital format; and conversely how
to interpret components/constructs within that digital format as
representations of components of a description set.
So for the case of the "Expressing Dublin Core in HTML/XHTML meta and
link elements", it describes how to encode a DC description set using
X/HTML elements and attributes, and how to interpret X/HTML elements and
attributes as a DC description set. Well, that's what it _should_ do!
But the current DCMI recommendation
http://dublincore.org/documents/dcq-html/
actually pre-dates the development of the DCAM, so that mapping between
the abstract information structure and things in an X/HTML document
really isn't as clear as it should be.
DCMI is working to address this by updating the "encoding guidelines"
specifications, and I've recently circulated a draft of an updated
version of the X/HTML encoding guidelines
http://dublincore.org/architecturewiki/DCXHTMLGuidelines/2007-07-27
which does describe the encoding on this basis. It's still very much
work-in-progress, but it might help answer some of the questions you
raise here.
I agree that the use of terminology can be problematic at times,
especially when sometimes the same term is used in different contexts to
refer to different concepts. Terms like "element" and "schema" are the
usual suspects! We've tried to be more consistent in recent documents,
and to be clear about the context, and when we are referring to e.g. XML
elements rather than DCMI elements, and so on.
But the collection of documents available on the DCMI Web site has been
developed over a considerable period of time, over which period both the
concepts and the terminology have evolved, so I'm fairly sure there are
cases where documents are inconsistent with each other, at least on some
of the details! :-(
> 3) What is DCTERMS.somethinghere ? What does the DCTERMS
> mean? do i need to use it within my document?
The use of these "prefixed names" (i.e. names of the form DC.xxxx and
DCTERMS.xxxx) is an abbreviation mechanism for URIs. i.e. all DCMI terms
are identified by URIs, and description sets refer to terms using those
URIs. The "encoding guidelines" for X/HTML specify that some URIs in a
description set can - in this syntax - be abbreviated using these
"prefixed names".
The reason for using two different prefixes is that some of the metadata
terms owned by DCMI have URIs beginning
http://purl.org/dc/elements/1.1/... (e.g.
http://purl.org/dc/elements/1.1/title ) and some of the metadata terms
owned by DCMI have URIs beginning http://purl.org/dc/terms/... (e.g.
http://purl.org/dc/terms/created ). The reasons for the assignmewnt of
those different URIs are historical and based on social/organisational,
rather than technical, factors, but that's the way it is. And so the
prefixed name abbreviation mechanism requires the use of two different
prefixes to represent URIs based on those different "roots".
(In fact there's a proposal undeer consideration at the moment to create
a new set of terms and URIs so that there will be a new set of terms
with URIs beginning http://purl.org/dc/terms/... Corresponding to the 15
existing properties with URIs beginning
http://purl.org/dc/elements/1.1/... So if that proposal is adopted, you
may find you can avoid the use of two different prefixes in X/HTML the
future)
> At the moment im only concerned with using the 'standard'
> elements within my documents, to add some extra value to them.
>
> I have read the documents, and find some of it confusing,
> especially the document that shows how to uses DC MetaData
> with HTML, i get easily confused by DC.element and
> DCTERMS.element, and dont understand the difference between them both.
>
> Thanks for reading my essay,
Pete
---
Pete Johnston
Technical Researcher, Eduserv Foundation
Web: http://www.eduserv.org.uk/foundation/people/petejohnston/
Weblog: http://efoundations.typepad.com/efoundations/
Email: [log in to unmask]
Tel: +44 (0)1225 474323
|