RE: Best citation format for accessibility

+1 to the mapping, but I also want to point out that re: the "something else" issue, the cited article is not part of the HTML page, but the _citation_ is part of the web page.

That's not necessarily to disagree with your point. One of the reasons I asked Robin if there was any inherent conflict—and this may not have been clear—was the prospect of using BOTH in the same HTML page. So is THAT a conflict?

The reason I might want to do that is specifically in order to provide ARIA for AT but to use the markup Robin suggested for its broader usefulness.

So I actually wasn't suggesting they were interchangeable ways of doing the same thing, necessarily.

From: Ivan Herman [mailto:ivan@w3.org]
Sent: Wednesday, September 23, 2015 12:47 AM
To: Robin Berjon
Cc: Bill Kasdorf; W3C Digital Publishing IG
Subject: Re: Best citation format for accessibility



On 22 Sep 2015, at 23:37 , Robin Berjon <robin@berjon.com<mailto:robin@berjon.com>> wrote:

On 22/09/2015 17:25 , Bill Kasdorf wrote:

Thanks. And do you see any inherent problem in this
microdata/RDFa-in-schema.org<http://RDFa-in-schema.org> approach conflicting in any way with
ARIA? Is there any reason the same sort of semantics couldn't or
shouldn't be captured both ways? This is not a rhetorical question.
Your answer may be yes, there is a reason, and if so I'd like to know
that!

I don't see a conflict in ARIA, in fact they could complement one
another nicely.

That said, I don't think that ARIA should become a universal semantics
layer. Capturing some important relatively high-level landmarks and
features in the way that the DPUB ARIA currently does makes (a lot of)
sense, disambiguating a page from a page range very likely isn't
(whereas it could be useful information for a machine).

I would go beyond that: I do not think ARIA is the right tool for this, conceptually.

As far as I am concerned, I make some difference between ARIA and RDFa/µdata. ARIA provides "semantics", ie, structural information, about the HTML page itself. It adds a semantic layer on top of the tag soup to describe what a tag is used for. RDFa/µdata, on the other hand, uses the HTML tags to provide semantics about something else, which is not, per se, part of the HTML page. Using URI-s for this purpose is one of the strengths of RDF(a). In this case, RDFa/µdata provides semantics about an article through its reference (DOI, HTTP URL, whatever). ARIA cannot, and shouldn't, do that.

It is a subtle difference, but an important one imho. In this respect, expressing structural semantics with ARIA is the right thing to do; but expressing bibliographical information of an article which is not part of the current document isn't, at least in my opinion.

B.t.w., I would be interested to see a comparison between the JATS and the BIBO terms. BIBO is an RDF vocabulary, and hence can be readily used (and is used in some places) to describe bibliographical items using, eg, RDFa. I have never looked at JATS but if we really feel it is necessary to represent that in terms of an RDFa vocabulary, the good thing would be to map it on top of BIBO and not to re-create yet another vocabulary (there are also at least 3 different mappings of BiBTeX to RDF, for example, all incomplete from a publisher's point of view; we should try to avoid this type of proliferation).

Ivan





Syntactically there is no conflict, so we ought to be fine.

--
• Robin Berjon - http://berjon.com/ - @robinberjon
• http://science.ai/ — intelligent science publishing
•


----
Ivan Herman, W3C
Digital Publishing Lead
Home: http://www.w3.org/People/Ivan/

mobile: +31-641044153
ORCID ID: http://orcid.org/0000-0003-0782-2704