Semantic Web disillusion
- The 7 (f)laws of the Semantic Web by Dan Zambonini
If you are a SW addict like I am myself, consider to read this article. It’s very mind-sobering. The last year or two we all had uneasy feeling that something was going wrong but nobody ever dared to analyze and understand what and why. The author of this article did.
When it comes to the Semantic Web, you might call me a disillusioned advocate. I’ve been dipping in and out of the technologies for the last 5 years or so, but am increasingly frustrated by the lack of any visible progress.
Really, why the technologies which were going to be “a next big thing” in the Web and computer world 5 years ago, are still mostly the Glass Bead Game of few geek enthusiasts and academicians?
Look – ordinary Javascript hack with separate downloading the parts of a document got a great hype – so much that someone believes it is a basis of “the next generation web”. A company which developed a search engine (well, really good one, but just a web search engine) – is going to be (sorry) the next Microsoft. A simple web-bookmarking service and a photo-sharing tool caused the whole new information theory.
And what about the “the web of data”, “the web of knowledge”, “the web for machines” – a brilliant vision which had been so fascinatingly described in famous SciAm article by TBL et al? Had we been defrauded? No – we trapped ourselves.
If I would be asked in 2001: what is “the Web 2.0″? – I’d answered with no doubts: “the Semantic one, of course”. Just at that time I and few my colleagues left the reliable and profitable web-design business for contributing to the future – to run a company for developing the semantic ontology-based application server (my SOFA project grows out there). We failed with it – just because the competitors provided the same (from customers point of view) solutions based on PHP+MySQL, ASP or even “out of the box” like with Zope. They lacked any “semantics”, interoperability etc. but was worth much lesser efforts and costs. We thought then that the time for such things was not come yet, but it seems the time is not going to change.
SOFA (Simple Ontology Framework API) which was aimed to bring the semantical data processing into Java – is used, in fact, by literally only a few enthusiasts like I am. A firm which supported Kowari – a superb open-source RDF database – was collapsed and finally sold to someone who think that new Kowari versions would be harmful for them. And take a look to Kowari popularity even in its best time – at average 300 downloads per month (by SourceForge.net statistics) – it’s really very little. As for example, eXist (an XML database – I especially took the similar software on SF.net but unrelated to SW) has 3,000 downloads p/m – 10 times greater and it is obviously not a kind of a popular mass software. Even Protégé – famous Protégé – maybe the only succesful end-user ontology software was in apathy for the whole last year, until releasing new version right yesterday (08/01).
Now let’s see how the standards are used. The first RDF spec was released in ancient 1997. The children who were born that year, are in the school long ago. But where is RDF? A short-lived version of RSS – the technology which pretty well managed without any RDF before and is replaced steadily with “non-RDF” Atom now. What else – FOAF? But tell me why it needed to be RDF-based – it might be well done with any XML or even those “microformats”. It seems the RDF in both cases is used not for its real purpose. And there is only two real-world application I can remember, and I cannot remember nothing for OWL.
I think it doesn’t mean that the technologies were or are bad. There is something important beyond the technologies and it seems going wrong from a very beginning. Maybe something beyond competences of technology people at all. Yeah, the technologies were hardly understandable by public and by business people, but it was not the worst thing – it’s not a public technology after all. Many worse was that the people who could bridge the technology with the business and society didn’t understand it. “We don’t need an ontology editor, we already have Visio, thanks!”
And to be fair, how could someone ever understand it if there was absolutely no explanations what it is it and how it works – except for a few dull articles with diagrams how an author related to her web-page. “Maybe it makes some sense though I have no idea how it’s going to change the world. Ok, but what’s next?” And to learn more, please click here…
The Semantic Web lost those people though they were necessary – only they were able to find the real ground for our weak hothouse plants could grow and bear fruits. We need to get them to our side – knowledge management people, information designers, business experts, librarians – everyone who knows how our technologies could be applicable in the real world and who can show their benefits for all other people.
A hopeful thing was mentioned in the comments to Dan’s article – the Semantic Wiki extension for MediaWiki. Who knows what we, the SW folk, did when Wikipedia had been started? Maybe writing new model-theoretic stuff? And think where would we be now if Wikipedia had this feature from beginning…
