Semantic Web Round Up 2005-03-17


I've written a few entries about the Semantic Web already, but since my deadline is nearing on my essay for the MSc. course I'm doing, I've started rounding up a few links that I think is worthwhile sharing as well.

That's a debate that seems to be getting more and more heated.
On the more artistic side Dan Cooney has an interesting interpretation (via Edward Vielmetti).

The Semantic Web is here is a great introductory presentation by Eric Miller (via hannes.kaywa.com)

At heart of this debate is the discussion on whether folksonomies or ontologies provides the most value in a distributed, uncontrollable media like the internet.

This also tends to translate into a debate on whether to use micro formats or RDF as the carrier of semantic information.

Some thoughts on RDF vs micro formats

A quick and dirty RDF tutorial (via Ebiquity blog at UMBC - Thanks!) is a good way to start if RDF and the semantic web is completely new to you.

Regardless of past experience with the semantic web I would also recommend Tantek's presentation The Elements of Meaningful XHTML which provides a great overview of Micro formats and how to convey as much semantic information as possible through the use of XHTML alone.

However, while I see micro formats and overloading XHTML as useful to some extent, it misses a lot of the potential of the Semantic Web by not making the semantics of the markup easily discoverable, for instance through RDF-S or OWL.

RDF without the nasty syntax

One way of getting that benefit while at the same time achieving much of the simplicity and bottom-up approach to the semantic web is RDF-A.

Instead of specifying individual micro formats, RDF-A is an attempt to provide a trivially simple way of attaching attributes and elements to an XHTML/XML document from which it is easy to derive RDF triples.

The benefit is that you can do all the fancy stuff that people are working towards doing with RDF, while at the same time getting most of the simplicity that Tantek and the guys at Technorati is going for with the micro formats.

You can have your XHTML and still get RDF too

Another approach to solving this problem is GRDDL:

A mechanism for using transformations (in XSLT in particular) to express the relationship between XHTML dialects and RDF in order to expose the data in these dialects to the Semantic Web. The mechanism extends straightforwardly to XML formats in general.

GRDDL would allow micro formats to live alongside RDF eating agents by letting the XHTML specify a transformation, for instance an XSL document that specifies how to transform the XHTML into RDF.

GRDDL could be used for RDF-A as well, obviating the need for an RDF processor to specifically know RDF-A or future alternative syntaxes as long as it knows how to apply the transformations.

See GRDDL as the glue that makes it possible for you to more or less ignore the "war" between RDF and micro formats as markup if all you want to do is write apps that consume RDF.

In the end I prefer RDF-A over Microformats because they seem to give more potential for reuse.

Mike Linksvayer knows this stuff better than I do, and has this to say (this was where I found cool stuff like RDF-A)

Who'll tag all this stuff?

In this article Russel Glass raises exactly that question and goes on to say:

Just as the Web, however, allowed an upstart like Amazon to compete with Barnes & Noble, the Semantic Web has the potential to level the playing field again for a whole new generation of startups. With the Semantic Web in place, any vendor will have the ability to tag their product information and make it as easily accessible as Amazon.

I agree with this. One of the key values of the semantic web is that it breaks down virtual monopolies. Today, it takes a tremendous effort to gather together product information to set up a product search, for instance, because the information is harder to find than need be. Contrast it to how easy it is to find news items via RSS. Now imagine that a retailer can achieve only a one percent sales increase thanks to aggregators and new search services if they tag their data. They'd jump on it instantly.

Will the Semantic Web become a success?

Notes from The Semantic Web: Promising Future or Utter Failure, a panel discussion with Linksvayer, Galbraith, Marlow, Haughey, and Champeon is a good read for a quick introduction to the various viewpoints.

philwilson.org: Finding related items in your RSS datastore points out another issue: Once you have all this information tagged, as we do with blog entries via RSS for instance, how the f**k do we actually find what we want in between all the cruft that's bound to show up (take a look at the list of blogs at blo.gs for instance, and you'll see a worrying number of pure spam blogs), as well as all the stuff posted with a good intention that simply isn't your cup of tea.

Listed below are links to weblogs that reference Semantic Web Round Up:

» Metadata and the Semantic Web from Denise LeVan's 718 Weblog
“RDF - the Resource Description Framework - is a foundation for processing metadata; it provides interoperability between applications that exchange machine-understandable information on the Web.” According to Tim Berners-Lee, in his article The Se... [Read More]


blog comments powered by Disqus