Impressions Scholarcast

For years almost I have in the back of my mind the directive ‘I must get more deeply into the semantic web’. However, the thought of linkbases and metabases, the mysterious RDF triples that are displayed 5 different ways, and vocabularies and ontologies in weird formats has always put me off.

You can imagine my excitement when listening to an ITConversations podcast entitled ‘Adding Semantics to HTML with RDFa’. Dubbed Semantic Annotation for (X)HTML, RDFa involves adding XHTML attributes, hence the ‘a’, containing simple values from which the normal content of RDF triples can be generated easily. No XHTML or XML elements are used so the usual web page publishing software and CSS styles sheets are not affected in any way. Imagine the simplicity of just adding attributes for the machine-readable semantic data to XHTML pages created as per normal.

The RDFa Primer which is a W3C Working Group Note envisages the following semantic transformation:

A cleverly-designed minimal set of attributes are introduced using existing attributes where possible:

• about – metadata resource (current page by default)
• rel and rev – relationship or reverse-relationship
• href, src and resource – partner resource
• property for element content
• content – where element content not accurate
• datatype – force data type
• typeof – for  RDF type(s)

There are several benefits but the main one are publisher independence in that each site can use its own standards, data is not duplicated (DRY, do not repeat yourself), and XHTML and RDF values appear in the same document but are easily separated with existing tools.

Of course you are encouraged to use existing vocabularies where they exist. An example using Dublin Core taken from the Wikipedia entry is:

Of course it is one thing to mark up the content, but display tools are needed to augment browsers and the like to make use of the semantic data in useful ways. The Semantic Radar Firefox extension alerts the user in the Firefox status bar to the presence of RDFa information in a page and giving access to the generated triples and RDF data:

It is to be expected that Creative Commons has already adopted RDFa so when you generate a licence and copy the resulting XHTML to your web site or blog it will contain RDFa attributes built in:

Digg is also using RDFa as can be seen by inspecting the XHTML of a Digg page but for some reason Semantic Radar does not find it.

Discovering RDFa has begun to change my attitude to the semantic web and I will now follow progress with anticipation.

Some fun via Peta:

Not sure about the prehistoric life, plants and animals though.