First RDF Schema for a Semantic Web enabled Drupal

As a semantic web researcher and developer, my goal is to bring these technologies to the
lay people. The main problem is the common chicken and egg dilemma,
where the semantic web technologies need semantic data to become truly
useful and powerful, but nobody wants to produce such data until they can see how powerful the semantic web is.

There is an immense amount of data available on the internet spread over millions of
HTML pages, PDF documents, et cetera. These formats have been designed for
making these documents understandable for people, but not for machines. In this instance RDF
comes in as a language to describe data and relationships within the data. From a web of documents we evolve to a web of pieces of
data, i.e. concepts, items, ideas, events, people, you name it. Each of
them can be identified by their own Uniform Resource Identifier (URI), and the web becomes a global database.

To achieve this challenge, data producers need to adapt their
applications so that they supply RDF data in a standard fashion which
will be understood on a global scale: it's the role of the W3C
to coordinate this effort. Drupal and other CMSs are
important data producers and as a result could set the standard for weaving the Web.

I'm glad that Dries, in his Keynote at Drupalcon Boston 2008, understood the potential behind the semantic web and took the decision to push this effort into Drupal. I'm thrilled to see this happening in Drupal 6 and in Drupal 7 core. For those interested, here is the semantic web audio part of Dries' keynote from the original audio recording. The full video of the keynote is also online.

An interesting mashup demo video was showed during the keynote, and it demonstrates what semantic web is capable of and is a good example of how it can be used in Drupal. When I first saw the video, I felt the explanation of the SPARQL query was confusing. Some lines are highlighted with the following explanation:

It can even do this across multiple datagraphs at the same time, which
means that creating complex queries across disconnected data sources
can be very simple.

While it's true that SPARQL can query multiple sources, the query presented here is not the right example. The highlighted lines are just namespace prefix definitions commonly used in SPARQL to make queries more concise and are taken from the Turtle format. A query across multiple data sources would look more like the following (note the NAMED and NAMED GRAPH lines):

PREFIX foaf: <http://xmlns.com/foaf/0.1/>
PREFIX dc: <http://purl.org/dc/elements/1.1/>
SELECT ?who ?g ?mbox
FROM <http://site1.org/dft.ttl>
FROM NAMED <http://site2.org/alice>
FROM NAMED <http://site3.org/bob>
WHERE
{
?g dc:publisher ?who .
GRAPH ?g { ?x foaf:mbox ?mbox }
}

RDF Schema for Drupal - Why bother?

As I emphasized in the beginning of this post, the basic principle to remember is data portability. It's important to make sure that the way you express your data is machine readable and understandable. Most platforms need to describe their users and the relations between them; the data these users posted, and how they interlink to each other. Describing what a piece of data is with taxonomy terms and how these terms are related is useful to many systems. These cases are not specific to Drupal, standards have been put in place in coordination with the W3C:

• Dublin Core (DC) fairly general vocabulary for describing online materials such as text, image, video, sounds or web pages.
• Friend Of A Friend (FOAF) for describing people, their activities and their relationships
• Semantically-Interlinked Online Communities (SIOC) for describing online communities and the content they produce on internet such as discussions, blogs, forums...
• Simple Knowledge Organization System (SKOS) useful for classifying knowledge and defining taxonomies

I proposed a first RDF schema for Drupal on groups.drupal.org, and it's great to see that some knowledgeable people from the Linking Open Data project such as Frédérick Giasson have already started to give their feedback! Let's see what we can build up on it.