Frédérick Giasson revealed a pre-release of the zitgist semantic web browser. It's a great leap forward in making RDF data readable and browsable to people who would be perfectly happy in their lives never needing to see the acronym RDF. (That said, I did hear tweets from a couple people who didn't find it quite as readable as might be desired--more on that below.)

The timing of the pre-release also relit the fire of working on ideas for how to capture and represent the intellectual life of the students at UMW. Okay, it was more like pouring gas on that fire. I'd been working on gathering up as much basic data about the people and courses at UMW as I could and putting it into RDF form, and the possibility of giving that data a test drive in the zitgist browser was a little too much for me to pass up.

(Speaking of fires, though, I fried the motherboard on my laptop over the weekend churning through lots of this data and spitting it out in a passable form. Thankfully, Barry in UMW's User Services hooked me up with a loaner and swapped out the hard drive for me. But I still fear trying to rewrite the data in another pass to clean it up more. Hopefully, what I have will get close to the ideas, if not a technically proper implementation of them.)

One essential ingredient to what we've been talking about is the idea of capturing the unexpected connections between readings, lectures, different courses, cocktail-party conversations, etc. that students and faculty come up with. So here's a possibility for what one suchconnection might look like, modeled in RDF, and viewed through the zitgist browser. As usual, I'm pressing Gardner's good graces by picking on him first. He's teaching a course this summer (starting today!) called "Memex to YouTube: New Media." So I pretended that I am a student of his (not too much of a stretch, come to think of it!), and thought, "Hmm...I wonder how close the zitgist browser is to the vision of the Memex." Not a fully formed idea, but perhaps a happy start, and it'll make an adequate example. A real-life input mechanism for it is still to be figured out, so I typed the data in directly. But, here's what the zitgist browser for that starting point looks like. (Warning, the web addresses might be off-putting and are, I think, a real issue. But bear with me and they should make sense).

I've given a description and a title to the idea. Conceivably, this could also be the topic of a blog post in which I'd think through the idea more, and a link to that post could also be shown here. Now, about those web addresses for 'creator' and the 'subjects'. The semantic web identifies everything with a web address (I'm fudging a bit on that, but I also promised that this post would be in English, so I won't elaborate). But what's more, that web address also points to information about the thing it identifies. So the web address for the creator points to information about me. The web address for the second subject identifies Gardner's course, and gives more information about it. So it's really just an under-the-hood view of:

,where everything in the second column is a link to more information. I'd have to agree, though, that it would do a lot for the readability for general users if things like real names and titles could replace the web addresses in the links. The choice to recreate the visual appearance of a tabbed notecard was a brilliant move toward making the visual appearance comfortable and familiar, but the appearance of raw web addresses undercuts that a bit.

So, lets follow a link. Here's what I get when I follow the link to Gardner's course. It's listed as a Group (partly) because I'm going to call an actual course a group of people. That's as opposed to the abstraction of the course designation. That is, "English 375PP" is a course designation, primarily for administrative bookkeeping. "Gardner's English 375PP, section 1, during summer 2007" is primarily a group of people, embodying actual practice and engagement in whatever "English 375PP" designates. Hence, I'm also calling it a "manifestationOf" "English 375PP". (There are conceptual wrinkles here I'm still wrestling with, though). And, we've got the same game with the web addresses. So, for example, the last line just says that the course is taught by Gardner--more info at that web address . . .

. . . and here's what's there: a list of courses he's teaching, and the department he's in. There's also a picture of Gardner, but I had to crop off the right side of the screenshot to make it fit in this post (sorry, Gardner). Similar browsing can be done with any of the links. That's what it's all about--browsing linked data.

Here's a link into the zitgist browser for this. Just one thing, though, if/as you look around in it--this is a rough draft of my data being viewed through a pre-release of the zitgist browser (as of Monday, June 25). It's pre-released so he can work out bugs before an official release, and so I'm a little nervous about my drafty data ending up giving a misrepresentation of his excellent work or, worse yet, suggesting that there are bugs in his app that are really bugs in my data-scraping and data-modeling. That's a fumbling way of saying that, as a pre-release, the main audience of the zitgist browser right now is probably people who spend a lot of time in the depths of RDF, and I hope that my exuberance with the app and what it can do with UMW data won't cause problems. I'd also recommend looking at the links into the data browser in the blog post about it (see th link at the top of this page.