Monday, June 04, 2007

Lots of Zepheira Press

Zepheira has been getting a lot of press recently. The Zepheira news page certainly has links to more podcasts and respectible publications than any other company I have have the privilege to be part of.

Investors Business Daily came out today with an interview of several key SemWeb leaders, including Eric Miller. The article itself expires after today (!) so I have created a PURL for it so we can eventually move the redirection when IBD pulls their head out.

The article was typical journalistic fare, attempting to be "balanced" in the same way that Fox News is. That is done by showing both sides of a story - even when there are not, in fact, two sides. We see this sort of nonsense with science reporting when reporters go increasingly out of their way to find a few scientists willing to publicly doubt the phenomenon of climate change or willing to say that nuclear fusion will happen commercially within the decade. In the case of the Semantic Web, the same old doubters of 1999 come out to play and get quoted every time, ignoring the clear commercial successes of the last few years.

It is interesting to see Tim O'Reilly routinely quoted in the press now putting the SemWeb in a more positive light than last year: "The Semantic Web is the idea of marking up computer information in such a way that computers can infer meaning from it."

The journalist managed to get basic facts wrong, too, as when he says, "Internet protocols make it hard to search for many basic items on a Web site, such as a simple address or phone number." or "A resource description framework (RDF) and Web ontology language (OWL) are new technologies that can solve that problem. They serve as a kind of wrapper or tag to describe the data inside." RDF and OWL are hardly new (both were standardized in 2004 and pretty stable years before) and of course the Internet protocols have nothing to do with linguistic searching.

He also gets some things right: "This new vocabulary lets computers find and access data on their own. The goal is letting the machines perform rote tasks to gather information and merge the results.", and he mentions products from Oracle and Adobe.

Eric is quoted reasonably. That is a relief. "These Web standards should help companies spot new relationships among huge sets of data and use the findings for better conclusions about their business, says Eric Miller, president of Web startup Zepheira."

I rather liked the MySpace example and wonder if it came from Eric:

For instance, MySpace might let personal pages share information with the pages of relevant friends or colleagues in the social network.

Take someone whose MySpace page describes a fondness for vintage jazz. By entering that information once, that person could automatically be linked to others who share the same interest.

Furthermore, that information could be applied to future Web searches for new music releases. In effect, using metadata could become a way to make MySpace "truly mine," said Miller.

"This means there is a much more flexible, personalized integration point to really connect people," he said. "The notion here is to enter data just once, but to use it often."

No comments:

Post a Comment