Raw

A good post from Dan Zambonini: Why the World is ready for the Semantic Web. He suggests that:

...some of the biggest and best projects on the web today are based on Semantic Web principles - they just don’t happen to be using Semantic Web technologies.

He lists various technologies (Google, RSS, Tagging, Social Networking, Mash-ups) and describes them in SemWeb terms. One tech I'd add to that list is microformats.

I reckon he's essentially right, and note the common aims of Web 2.0 and Semantic Web efforts: Web as Platform, Web of Data.

There's also the dark web tech to consider, like having sites based on SQL DBs and only exposing a human-readable HTML view of the data. The barrier to exposing this material on the Semantic Web (and thus increasing the surface area of the publisher) is getting lower all the time - D2R is just one example of the various toolkits around that can enable this. Now SPARQL is approaching Rec. status, it's likely that a lot more applications for offering SemWeb-queryable endpoints will appear.

Then there is the point that the technologies listed above have their primary adoption in the general public, lay user demographic. There's also pretty hefty amount of specialist stuff using SemWeb tech, notably around the Life Sciences.

But currently the greatest weakness of things like RSS, tagging and most social network efforts is that they tend to be hard to integrate with each other cross-domain. The apps tend to be built as vertical, domain-specific "stovepipes". As Dan points out, all of these can be built using SemWeb tech, and integrateability (?) essentially comes free. 

Toolkits are now available for every language under the sun, leaving the greatest barrier to adoption the level of familiarity of SemWeb tech amongst developers. Aside from general awareness, there are plenty of common misconceptions. Again SPARQL is potentially a powerful hook, compared to any web tech it's relatively simple, and offers a gentle slope to understanding the RDF model. Although I'm still rambling on, there is a lot more showing going on these days, and SPARQL (with JSON results) with Ajax makes a very compelling compelling combination.

At the Jena conf. one of the HP guys (sorry, I forget who) pointed out that the growth of Web Services (both WS-* and REST) has done a lot to reveal just how hard integration can be - when you don't have an appropriate solution available.

I personally believe we will see significant growth in the adoption of SemWeb tech over the next few years, largely because of that integration bottleneck. You can solve such problems using relational DBs, XML tech and/or things like Linq, but RDF renders many of the otherwise hard parts trivial. It supports both traditional and (especially) agile methodologies. The company that makes use of SemWeb tech has a head start over less (web-) systematic approaches. In a huge, diverse, heterogenous environment like the Web presumably Darwinian natural selection is a factor in adoption. As more systems appear on the Web with SemWeb capabilities, and the more data appears in a machine-friendly form on the Web, the more the network effect will amplify the benefits.

One final point, Semantic Web technologies already offer a relatively low-cost mechanism for integration of a lot existing Web data. RSS/Atom can be interpreted as RDF. So can most tagging and social network data. So can microformats. Google's search is the odd one out, but they have things like GData which can be interfaced with the SebWeb, and integration with search will no doubt happen within Google's wall.

Web 2.0 is good for the Web. Because the Semantic Web is an extension of the existing Web, what's good for the current Web is good for the Semantic Web, and vice versa. Whether or not the world at large is ready for it is another matter (was it ready for the Web?), but I reckon the Semantic Web snowball is already rolling. Ready or not, here it comes.

 

@en