RealTime IT News

Semantic Web to Take Center Stage at WWW2004

Discussion about what may be the most exciting and controversial concept in Web development will be the main course on the menu at the 13th annual World Wide Web Conference in New York next week, according to event chairman Dr. Stuart Feldman.

In offering a taste of what attendees can expect, Feldman, also head of IBM's Internet Technology division, has told internetnews.com that the notion of the "Semantic Web," which some computing experts argue has ties to artificial intelligence, will be the talk of the town at the event.

"The Semantic Web is an increasingly hot topic," said Feldman, who began his career as a member of the original development team that built the UNIX operating system at AT&T's Bell Labs. "There's no question that the idea of being able to know what you're looking at and interpret it properly is about to come of age."

Pioneered by universally acknowledged Web creator Tim Berners-Lee, the Semantic Web is a concept that argues for a more intelligent Internet experience, where applications on computers are written to be more intuitive and accurate in processing data and finding results for end-users. Berners-Lee, himself, has described it as a "giant database."

The idea behind the Semantic Web is to give data more meaning through the use of metadata, which describes how, when and by whom a particular set of data was collected, and how that data is formatted.

By adding metadata to the current Web, the Semantic Web can allow people and machines to make use of data in ways that previously haven't been possible.

There are two main components to the Semantic Web: Resource Definition Framework (RDF) and Web Ontology Language (OWL). Both were ratified as standards by the World Wide Web Consortium, a group devoted to hashing out Internet rules, which is led by Berners-Lee.

"It's a very impressive push," said Feldman, who has worked with Berners-Lee on various Internet-related projects. "RDF and OWL have gelled very well. One more or less tells you how to find and utilize information [RDF]; the other how to understand it and its taxonomies [OWL]. These topics hark back many decades in the artificial intelligence literature and such. What's important is that the current definitions and standards are computationally useful and people are finally starting to use them."

Feldman said a fair corollary is looking at Semantic Web, still in its infancy, the same way developers looked at XML in 1997 or 1998. At that point, most software programmers could scarcely imagine that the language would be used to signify content on the Web.

Now, it underpins the Web, and serves as the backbone for many applications written today, particularly for next generation software paradigms such as Web services and new service-oriented architectures .

If the Semantic Web follows the same path, it could change both the way people program and use the Web.

"At a certain point, people just stopped arguing about XML," Feldman said. "The reality is it's good enough and people more or less decided it wasn't worth doing another iteration and fighting it. I think that is a sign that the field is maturing. I think it's possible the Semantic Web standards are in the category of 'We can argue forever, so why don't we just use them.'"

Feldman also addressed the Semantic Web's controversial link to artificial intelligence (AI), borne of the fact that the specification uses the word ontology, which is a specification of a conceptualization at its base.

"You can reasonably argue that the Semantic Web is the realization of some of the dreams of distributed AI," Feldman said. "The proponents carefully avoid using 'AI' in some contexts because of the unfortunate PR that goes with the word."

Feldman refused to discuss on what plans IBM has for Semantic Web, acknowledging only that some developers in his own group have used RDF to give presentations and that the Armonk, N.Y., company is looking at how the technology affects the company's information moving forward. He did, however, acknowledge that a combination of XML, RDF and Web services could be a possibility.

Some vendors, on the other hand, are outwardly employing it. Adobe has based its Adobe XMP (Extensible Metadata Platform) on RDF. The Adobe Creative Suite lets users create assets powered by XMP that can be repurposed and consumed more effectively across several different media and domains.

For example, users can take a picture and crop it without fear of losing the original, which can be used for other publications. Metadata has become equal to, or more useful than, the applications that created it.

Meanwhile, Sun is employing RDF for an in-house digital asset management system to manage content in multiple languages and departments and stitch them together.

As for other topics at WWW2004, Feldman said classic problems of security, and the engineering of human interfaces, such as multimedia, will be fodder for discussion, as well as other "brutally nerd-o-cratic topics." The engineer said there would be roughly 75 or more talks on technical issues, whittled down from three to five times as many submissions.

Tuesday will be a day for topical tracks for less technical users, such as "Will blogging change the function of news gathering?" and "What is the future of publication?"

Berners-Lee, himself, will kick off the keynotes, which will span the rest of the event, on Wednesday. He will be followed by Dr. James J. Duderstadt, President Emeritus and University Professor of Science and Engineering at the University of Michigan; Udi Manber, CEO of A9.com, a subsidiary of Amazon.com that develops technologies for e-commerce search; Rick Rashid, who oversees Microsoft Research's worldwide operations; and Mozelle Thompson, Commissioner of the Federal Trade Commission, who has been at the forefront of the war against spam, online scams and identity theft.