November 11, 2003
Semantic web redux
Semantic Web activity, alright. This week's hot topic in blogland is the semantic web, which has to be a good thing - focusing attention on it may allow us to make some sort of progress. It all started with Clay Shirky's article on The Semantic Web, Syllogism, and Worldview. This was an interesting article (although I don't agree with much of it), and Shirky concludes: "much of the proposed value of the Semantic Web is coming, but it is not coming because of the Semantic Web".
Shirky's article certainly stimulated a lot of debate. Tim Bray's rebuttal, as always, hits the nail on the head: the idea that data.somecompany.com could be an XML representation (maybe in the Extensible Business Reporting Language, XBRL) of all the business data of a company. This would be a very good first step.
The Semantic Web is something I've spent a lot of time ranting about, in association with the idea that we must unleash dark data - making the incredibly useful back-office databases accessible via the web (and preferably augmented by machine-friendly formats, for example OAI).
I don't think any of us will get there by trying to tackle the big picture in one go. It'll take baby steps. And I propose Cocoon as your companion while you take those steps, as it's like a guide on how to walk. By building multichanneling, xml versions, RDF, and maybe even support for web services into your sites, you will suddenly find that the semantic web is a reality without noticing that you've built it.
Danny Ayers weighs in with his own rebuttal, concluding: "there are systems based on Semantic Web technologies working now. The web is working now". This is more or less the response I gave on the JISCMail web-support mailing list last month, when I started flaming a "there's no point in xhtml /w3c sucks / semantic web is nonsense" troll.
Dan Brickley, who I had the pleasure of meeting last year in Boston and who works with one of our customers, gave a very measured and informative response. In it he points out that there's more to the Semantic Web than ontologies and reasoning engines.
Marc Canter is much less forgiving of Shirky's essay: "Rolling up your sleeves, digging into one area [...] and working through the ugly details - is what it's all about. [...] The technologies that surround the Semantic web [...] are here today, while the dream of the semantic web is still years away.� But unless we start building NOW [...]". A pragmatic approach I fully agree with.
Other commentary on the issue comes from David Weinberger (who's not sure whether it's Semantic Web or semantic web), fellow Cocooner Guido Casper (who picks up on the potential of data.somecompany.com), Kevin Werbach (who doesn't get that Shirky's argument is a straw man, and supports him), Dave Winer (no surprises, Dave is typically more excited about the opportunity to be controversial than to analyse the argument), and Burningbird, who really gets it: "it will come about when people see the benefits of smarter data. Small pieces, intelligently joined". mamamusings is collecting a more comprehensive list of responses.
Ultimately there's no quick and simple solution to the Semantic Web. It'll take many more blogfests, rants and heated debates before we really figure out what's going on. In the meantime, we are quietly building the elements of the Semantic Web around us, and it may just happen sooner than we all think.
Posted by savs at November 11, 2003 9:22 PM