Yesterday was spent drinking lots of coffee with a roomful of hackers working on cultural data, and putting together interesting things in the space of a single day. There was a prize in it for teams that built apps using data from Kasabi and were available outside of the room. It’s a good way to spend a Wednesday, and I’m very grateful to everyone who came along and built things with Kasabi!
The room soon broke into groups focused around ideas for apps, and I was able to move around and see what everyone was planning. As a bit of an experiment, we were joined on a Google+ Hangout by Daniel Koller, who had the distinction of being a disembodied head carried around to each planning table.
By the end of the afternoon, we had a set three hacks which qualified for the prize.
Alf Eaton, Andrew Nicolaou, and Dan Nuttall worked along with Leigh Dodds on something to bring people-tagging to pieces of art. Using the Government Art Collection dataset, some magic, and JQuery, they put together this bookmarklet. When viewing a piece of Government Collection artwork (like this one) or a painting from Artfinder, you can annotate it with the name of the people it portrays. When the bookmarklet is clicked, the hack places the annotation box around the area it guesses is most likely a person’s face, and makes some suggestions about who is in the work. Behind the scenes, the hack is both reading and writing data, storing the information from each annotation in Kasabi.
For the second hack (which you can see here), I’ll let Michael Lenahan (who worked with Michael Sadler, Rick Yagodich, Betony Taylor and Benjamin Nowack) explain about what they did from the blog post he wrote up during the event:
These maps show the home towns of bands listed on wikipedia/dbpedia. See these examples for Germany, United Kingdom and Canada. To see another country, replace Germany, United_Kingdom or Canada at the end of the url with the country name you want.
Benji created an interface for the maps here. The maps have varying amounts of information for each country, mainly due to some gaps in the dbpedia dataset. Think of it as a prototype, people!
So, how does this all work?
The starting point is the dbpedia data set, which contains the information stored on wikipedia.
Consider this page about Kraftwerk: http://en.wikipedia.org/wiki/Kraftwerk.
The dbpedia resource for Kraftwerk is here: http://dbpedia.org/page/Kraftwerk.
On that page, search for “dbpedia-owl:hometown”. We discover that Kraftwerk’s home town is Düsseldorf. Of course some Kraftwerk officianados know this already, but it’s a nice discovery for the rest of us.
On the Düsseldorf page we see that the geo:lat and geo:long values are supplied, so now we have latitude and longitude for Kraftwerk’s home town. Sweet!
The final hack came from Geoffrey Makstutis, Stuart Chalmers, Owen Stephens and Julian Cheal. Using the SPARQL Views module, the team setup a Drupal site which acts as a kind of browser for Government Art Collection pieces. You can see the hack here, which is a good example of pulling in data from another place, and displaying it in your own setup.
I’ll try to get some video from the day linked here in the next couple days, letting the teams present their own ideas, too.
I’d like to again thank everyone who came along, and those who dropped in virtually via Google, IRC and twitter. We’ll be doing more hacking soon, and if you would like to get involved with the developers’ side of Kasabi, you can drop me a line on firstname.lastname@example.org, and join the developers’ list for discussion and updates.