There’s a new working group for RDFa at theW3C – its mission isn’t radical, as the recommendation already exists, but its plans should address some of the concerns around the spec as it relates to HTML and wider adoption by web developers.

“One of the working group’s main goals is to add additional features, but not in terms of additional functionality,” says Ivan Herman, Semantic Web activity lead at the W3C. That includes “more features that simplify the creation of HTML files with RDFa. This is clearly one area.” That goes hand in hand with working with the HTML Working Group on incorporating RDFa in HTML5 and XHTML5, advancing it beyond its original definition for XHTML1.1. On the map is a version of the recommendation that is defined in a way that any XML application can use it. Herman points as examples to synchronizing the RDFa usage for the Open Document format that is used by OpenOffice, and to Scalable Vector Graphics (SVG), a language for describing two-dimensional graphics and graphical applications in XML, which allows adding RDFa attributes to SVG files but needs a more clearly defined way of doing so. Additionally, the group is going to define an RFDa API, a set of functions that could, for example, let Javascript accept RDFa functionality directly.

Herman expects that with the creation of this RDFa Working Group and some of the plans it has around the development of RDFa, the issues around RDFa will disappear. “We have plans for the simplification of RDFa,” he says. “We realize some of the features in the current RDFa make things a little bit too complicated, though I don’t think they are. But there’s the question of perception and some of the goals we have are going [to address] that direction. The only thing I hope is that there will be a proper professional discussion around that and that we will find the right compromises.”

Why’s it important to keep RDFa on a forward track? “RDFa combines two different worlds, the traditional HTML document-oriented web and the data-oriented semantic web,” he says. “You could look at RDFa as being another serialization that allows RDF terms to be expressed as part of an HTML file. This is very much the RDF angle. The other angle, which is more HTML-oriented, is that you have HTML files, and we want to add additional structure into those HTML files, additional metadata or whatever. You want this type of additional structure to be expressible easily in RDF in some way or other because then you can integrate it with all kinds of other data that is represented in all kinds of other serializations [different types of syntaxes].”

And momentum for it includes major search engines’ support of RDFa to Drupal 7 shipping with RDFa enabled by default on nodes, comments, terms and users. That one is huge, says Herman. He doesn’t claim to be a Drupal expert but says that “essentially what happens is if an organization uses Drupal for their web site, they will essentially produce RDFa as part of the web site they produce without even knowing it, let alone working on it really.” Herman also points to the U.S. Library of Congress Authorities & Vocabularies service that enables both humans and machines to programmatically access authority data at the Library of Congress via URIs. As an example he says you can head over here,
the URI for the term "Semantic Web," and it goes to an HTML page -- but this page also has its structure encoded in RDFa which can then be extracted by various tools.

There’s also, of course, the U.K’s e-government effort, which plans to make use of RDFa. “So you get a fairly large amount of data already today which is published in RDFa,” Herman says. “This cannot be ignored.”