Wednesday, May 19, 2010

University Websites, Linked Data and Giants

Finding out about what your member of Parliament is up to takes work - but web sites such as openparliament.ca have been wonderfully built and elegantly designed to make this work much, much easier.

Our library's web team has been recently activated and as chair, I've been casting around ideas on how we might re-build the foundation of our website as we move from a Lotus Notes Domino environment to a Drupal one. And the question that I keep coming back to is how can our website be more like openparliament.ca?

And so I've been re-imaging The University as Parliament. Parliament has bills, laws, committees, parties, departments, MPs, transcripts and media coverage. The University has courses, departments, committees, instructors, transcripts, research articles, and media coverage. Hmmm.. sounds somewhat similar.

So imagine your university website with a front page like openparliament.ca with the day's research and learning topics filling the page. Each topic would be associated with a class or research paper. Each paper or class would be associated with a researcher or instructor who would in tun be linked to all their various campus affiliations and the courses they are presently teaching. Each course would be associated with matching library resources...

If you had build a university website from scratch, it would be mad not to build linked data into its foundation.

But the trouble is, once I start browsing the Guides and Tutorials about Linked Data, I become so overwhelmed with the terminology of semantic linking and frightened of the enormity of the task before me that I promptly flee in terror.

So I am at a loss. I can see a conceptual structure that we can build our library website upon. I can see the opportunity to show others on campus (and on other campuses) what can be possible with linked, open data. Except I would much rather find a giant who's shoulders I can stand on so I see how this may be done.

But I have no giant. All I have is this handful of beans.

6 comments:

Lisa said...

you will BE THE GIANT. the rest of us will tremble at your feet.

Mita said...

You might like this slidedeck I just found on open data as its got a nice revolutionary political bent to it:

I love this: "What does Open Data have to do with failure? It changes (sic) the incentives from big, slow, deniable failures to small, fast failures from which you (and other people) can learn.

Mita said...

Now that my goal is public failure, I feel much better about this potential project...

Lisa Goddard said...

Hey Mita,

Drupal 7 will natively support RDFa, and there are bunches of tools out there to help produce more structured data on the back end, and to crosswalk library data to RDF. I just did a presentation on this at emtacl - video and slides are linked from the program, in case you're interested:

http://www.ntnu.no/ub/emtacl/?programme

Mita said...

Much thanks for the heads up on your presentation. I've been thinking about using Freebase as a means to get started on this project.

Please let me know if you think there is a better road to get to the promised land!

jonvoss said...

Lisa, that's a great presentation. Mita, as I said on Twitter, you have a great vision for what a university site might look like and do, which is the exciting part. I think you can start to build backwards from here, and it may be a long time coming, but by designing the platform for it with that vision in mind, it's well within reach.

It seems like for #hackacad, you would want to focus on the vision, because right now we have a chicken or the egg problem. It's hard to get people behind RDF publishing without being able to prove the use for Linked Data. A lot of the visualization and applications that would utilize Linked Data are not quite there yet.

But for the library, the more you can publish RDF from the get go, the better off you'll be. I'd love to hear more about this and offer whatever help I can to get you moving in the right direction.

Jon
@lookbackmaps