News

W3C Fires Up the Semantic Web With GRDDL

• By Kurt Mackie
• September 14, 2007

The Semantic Web effort aims to take the human-readable part of the Web and make it understandable by machines. One of the potential uses of the Semantic Web is to better enable mashups of data, which might, for instance, be used to improve data exchange for applications such as social networking, online shopping and scientific publishing.

GRDDL's role in all of this is to take the various metadata expressions coded in XML in a document and enable them to be transformed (typically by XSLT) into a common parlance.

GRRDL uses namespaces and compatibility with the Resource Descriptive Framework (RDF) to link to algorithms for extracting the data in a document. GRDDL's task can be facilitated by the use of microformats, which are proposed descriptors for commonly used data item (such as "name," "company," etc.).

One group involved in establishing common metadata labels is the Dublin Core Metadata Initiative (DCMI), which welcomed the W3C's GRDDL recommendation. The DCMI currently "is creating GRDDL transforms that expose Dublin Core metadata expressed in XML and HTML to the Semantic Web," according to a prepared statement by DCMI's Mikael Nilsson and Thomas Baker.

In the W3C's GRDDL announcement, Internet pioneer and W3C Director Tim Berners-Lee likened GRDDL to cascading style sheets in terms of their effect.

"Just as stylesheets make Web pages more readable to people, GRDDL makes Web pages, microformat tags, XML documents, and data more readable to Semantic Web applications, opening more data to new possibilities and creative reuse," he stated.

In addition to the GRDDL spec, the W3C also issued a recommendation on GRDDL Test Cases. The Test Cases are devised to check how software agents extract RDF from XML documents. The sort of agents that might be used for these tests include a SPARQL query service or a Web browser, according to W3C's Test Case document.