CNET también está disponible en español.

Ir a español

Don't show this again

Tech Industry

For W3C, it's a question of semantics

The influential Web standards group upgrades its Semantic Web activity, a push to make the Web a little more intelligent. But don't call it AI.

In its continuing effort to make the Web more intelligent, the medium's leading standards group has published a series of drafts relevant to its Semantic Web activity. But don't call it artificial intelligence.

The World Wide Web Consortium (W3C) last week updated several documents having to do with the Resource Description Framework (RDF) and the Web Ontology Language (OWL). Together, the protocols are supposed to make it possible for computers to communicate more context about information they send and receive over the Web.

The point of the Semantic Web is to remove the ambiguity that characterizes much digital information that now results in hit-and-miss search queries and the inability of different systems to discern meaning behind words and symbols.

For example, the RDF could help one computer system that recognizes the term "purchase order" to communicate with another system that uses the term "PO" for the same thing. It could also help computers associate two different but related terms such as "book" and "manual."

Some industry analysts and technologists criticize the W3C's Semantic Web activity as a pie-in-the-sky throwback to artificial intelligence schemes that has delayed more pressing work on Web services among other areas.

Perhaps for that reason, W3C members chafe at suggestions that the Semantic Web is reheated AI.

"It's not artificial, and it's not intelligent," said Eric Miller, activity lead for the W3C's Semantic Web Activity. "The conceptual models behind RDF are predicated on work in the digital library community. You can think of this as a common framework that supports thesaurus, taxonomies and classification schemes."

But critics continue to call the Semantic Web's mission a long-term proposition.

"It has been a challenge for the AI people forever to get the computer to understand meaning," said Ron Schmelzer, a senior analyst at ZapThink, a research firm that focuses on XML (Extensible Markup Language) and Web services. The W3C "is going after that same problem. This is something that's five to 10 years off. I don't want to be pessimistic, but it's going to take more than academics and the W3C to do this. It needs to be driven by industry and to be more focused on developing something specifically for commerce."

The first RDF document, , is an overview and tutorial of RDF.

The second is RDF Test Cases, a validation service that lets RDF authors check their work.

lays down the mathematical expressions by which RDF vocabulary terms relate to one another. RDF Schema says how RDF describes RDF vocabularies.

The Web Ontology Language, built on top of RDF, establishes additional capabilities and constraints for describing data.