Semantic Web holds promise for KM

"RDF is an ideal choice for integrating data from multiple applications having different schemas," says Xavier Lopez, director of product management at Oracle. "The RDF structure manages this information with all its rich relationship detail intact."

In much the same way that service-oriented architecture (SOA) has provided a loosely coupled way of linking business processes, Semantic Web standards offer flexibility in managing the data behind those processes.

"The standards do not go as far as defining a vocabulary that a specific community or organization would use," says Eric Miller, CEO of the Kadomo Group. "RDF and OWL rather provide a framework where different business domains can define their own vocabularies relating terms and concepts as different business needs require."

Miller served as the Semantic Web activity lead for the W3C and associate director of the Dublin Core Metadata Initiative, and has now founded Kadomo to help organizations deal with the practical aspects of implementing Semantic Web applications.

"Semantic technology is about wrapping content and presenting it again as data that can be combined with other data," Miller says.

Early adopters of semantic and ontology management tend to be larger enterprises or government agencies that have critical knowledge management challenges.

"Organizations involved in national defense, intelligence and homeland security agencies are deploying these leading-edge technologies," Lopez notes, "as are organizations in life sciences and the financial industry." Initial prototypes and applications have been behind the firewall rather than on public Web sites. Eventually, though, the data from some of those applications will be published and made available more broadly.

Some unexpected benefits have accrued to companies that have taken the first steps into this process, according to Miller. "Serendipitous interoperability and other side benefits have been greater than expected," he observes. In addition, some companies that never developed knowledge management programs have been able to leapfrog over earlier methods of pulling together information.

"We have seen some organizations use Semantic Web technology to integrate information when they did not have a way to do so before," says Miller. "In other cases, organizations wanted to weave together, in a graceful way, applications that have served them well in knowledge management but are not as flexible as they would like."

Progress has been slow, considering that the concept was first articulated about a decade ago by Tim Berners-Lee (who developed the first protocols for the Web and is director of the W3C). But adoption of any new technology does not proceed along a smooth and consistent path, and the potential rewards are considerable.

"We are looking at the possibility of accelerating progress on some of the world's most intractable problems," Miller says. "Diseases such as Alzheimer's and cancer would benefit immensely by automating even a part of the data analysis carried out by multiple researchers, and the Semantic Web can be a part of the solution."