Digital Web Magazine

The web professional's online magazine of choice.

RDF For The Rest Of Us : Comments

By Keith Alexander

July 30, 2007

Comments

Rik

July 31, 2007 6:09 AM

Well, I am a beginning web programmer but I totally agree on this one. Nice article and a nice read. Going to look for more articles from you :)

GRTZ Rik

C

July 31, 2007 12:12 PM

Nice idea, but for a real world sharing of data, RDF won’t cut it as it means 3rd parties will have to integrate a new type of ‘feed’. XML is currently the preferred method and is likely to stay that way.

Keith Alexander

July 31, 2007 2:37 PM

C, as I tried to explain in the article, RDF is a model, with many possible serialisation formats, including several in XML.

By itself, an XML format only provides a means of exchanging data, not of integrating it. RDF provides a common model for data, lowering the costs of integrating data from multiple publishers.

If you are already publishing XML, a relatively easy way to publish RDF as well is to use GRDDL – a standard for specifying an optional transformation of your XML format to RDF.

frances

August 3, 2007 8:27 AM

Fantastic article, Keith. It’s great to read a coherent and informative article on RDF since so much of the information out there is dense and technical. With the enormous amount of information going up on the web every day, I think RDF will soon become an advantage and then a necessity. I look forward to your next article!

ritzenhoff

August 5, 2007 12:04 PM

sorry, but i still don’t get it: what is rdf? another microformats standard? a meta level of html? or is it just like an rss feed? i can’t get the difference…

Keith Alexander

August 5, 2007 2:37 PM

Hi ritzenhoff,

I’m sorry that the article still left you confused. Perhaps the eRDF tutorial section was misleading as to what RDF actually is. (eRDF is just one of many possible ways of publishing RDF).

RDF is a model, a way of structuring data – like the relational model used by relational databases. If you want to publish the data in your relational database, you can export the data as a SQL dump, or CSV, or maybe XML (for instance). But if I want to use your data together with my data, it is a lot of work to merge the two databases, because they may well have different structures and semantics, and the primary keys from each database are unlikely to be globally unique.

This is the problem that RDF addresses; it helps us share and combine our data by providing a common structure (the triple), and a system of globally unique identifiers (URIs). It also gives us a means to define and share the terms that we use to describe our data. So, if you publish your data as RDF, I can easily save it to my RDF datastore, along with any other RDF data, and query it as a whole.

The confusing thing, perhaps, is that there are many forms in which you might choose to publish your RDF – and though they look different, they are all RDF. The canonical syntax is an XML format called RDF/XML; there is also the more readable, plain-text N3. eRDF and RDFa are two possible ways of expressing RDF within HTML, much like hCard is a way of expressing vCard within HTML (eRDF uses only existing HTML syntax, RDFa extends HTML with new attributes). (Don’t take this to mean, though, that RDF has a special relationship with HTML – it isn’t a ‘meta level of HTML’ )

Please let me know if this explanation helps.

Yours,

Keith

Levi Hackwith

August 7, 2007 10:05 AM

I enjoyed this article and the advantages make sense but is there any way you could show it to us “in action”? For example, show how flexible eRDF makes your data by feeding it into another site or feeding another site’s eRDF data into yours or translate an eRDF article into a PDF file. Is that a possibility or could you link to a site that demos these kinds of things? Again, awesome article and I love how well it was all explained (I’ve found RDF to be very confusing) but I’d love to see more “concrete” examples.

Keith Alexander

August 7, 2007 12:56 PM

Hi Levi,

A rather homely example would be
a blog post I wrote about RESTful PHP frameworks
.
The post is marked up with eRDF, and I used the DOAP vocabulary to describe the software projects the post is about.

The page was picked up by Ping The Semantic Web (pingthesemanticweb.com), (after being parsed into RDF/XML by triplr.org) where it was found, and the software-related data extracted and re-published by doapstore.org – a directory of software projects.

Now, that could still have been done if I had published with an XML format or a microformat specifically for software projects. But what I can do with RDF that I can’t do with specific formats, is relate lots of different things together.

The eRDF on my blog post, for example, not only describes those software projects, it also describes the blog post, and that the post is about those software projects.

If you want to get a better feel for it all, what you could do is find some RDF data (there are some links at the bottom of the article) and load it into an RDF browser (eg: http://demo.openlinksw.com/DAV/JS/rdfbrowser/index.html). It also helps immensely if you have a go at writing your own RDF (and I think eRDF is an easy way to start), and play about with it. Simile’s Exhibit, Piggybank, and Solvent are also good toys to play with in getting to grips with RDF.

Cheers,

Keith

Pars Narko

December 11, 2007 1:42 PM

I’m sorry that the article still left you confused. Perhaps the eRDF tutorial section was misleading as to what RDF actually is. (eRDF is just one of many possible ways of publishing RDF).

Tiff Fehr

February 18, 2008 6:03 PM

I’m closing comments on this thread, due to the amount of spam it is attracting. But I’m sure the conversation is still alive over on Mr. Alexander’s site, so please feel free to continue following RDF models beyond this primer.

Sorry, comments are closed.

Media Temple

via Ad Packs