Making it meaningful

If you're searching for a company's address on the internet, there's a good chance your chosen search engine will be able to find it by matching documents on the website containing the word 'address'. However, if that address information is 'tagged' as an address, search engines and directories can search specifically for that address data, or be sure that the content is indeed an address.

Being able to structure the information on your website in this manner will ultimately help your users or customers find what they are looking for, or differenciate your content from your competitors. This inference of meaning is what drives the semantic web.

Interest in the semantic web has gradually been spreading across the internet development community from a core of web evangelists and commentators to IA designers and developers building the next generation of websites and applications.

If you're new to the concept of the semantic web, it's essentially a web of data, readable by, and meaningful to machines rather than humans. In terms of the internet that we all know and love, it is being created by adding extra attributes and references to normal webpages so that humans or computers reading the page can understand the meaning and context of the content. Google's example is that of a page containing a recipe. Although Google might be able to guess that a page contains a recipe by its text content, semantic markup allows it to be sure that a certain section is an ingredients list.

Whilst many understand Tim Berners-Lee's vision for a web of data, evolved from human readable pages to machine readable ones , putting the concepts into practice can be a daunting prospect. The differing HTML5 and XHTML2 W3C specifications, as well as the somewhat competing standards of RDAa, microdata and microformats can be enough to put lots of developers off. As with the release of Betamax and VHS, Blu-Ray and HD-DVD, none of the different standards for semantic markup in webpages have yet become de facto.

For those of us staying the course, designing for accessibility and search engine performance provides a good stepping-stone towards building semantic web content: it helps us think about the content we place in our templates as information, rather than purely what appears before our eyes in the browser. If we're comfortable with this , the next logical step then is structured or semantic mark-up.

One of our current projects is for a fine art gallery and we're using RDFa, a semantic format supported by the W3C. The Resource Description Framework (RDF) specification has been around for over a decade but it is only relatively recently that it has been possible to add this type of data to a normal HTML webpage. We've chosen the RDF format as it is also used by the Visual Resources Association, who provide a data standard that we can use to mark-up the pages that contain artwork.

Modifying the templates to include the RDF attributes has been pretty easy but we won't know how well the extra semantic data performs until the site is alive on the web. Obviously, we hope it will help search engines find the data and in turn, help users find the content they are looking for. Alternatively, it could have other benefits to art institutions or students using the data for their own research. Given the infancy of the format and its use in search, it may only have a relatively minor effect.

Either way, it's been a rewarding experience, both in terms of exploring the broader concepts of semantic markup and the specific standards used in the classification and description of art.