"Any medium powerful enough to extend man's reach is powerful enough to topple his world. To get the medium's magic to work for one's aims rather than against them is to attain literacy."
-- Alan Kay, "Computer Software", Scientific American, September 1984
Zotero, OpenLibrary, and Some More Components of a Giant EduGraph
0Yesterday's announcement that Zotero 1.5 beta service is up is a good prompt to some more thinking about components of a Giant EduGraph. The Zotero service allows social network features like following other users to get updates about items added to their libraries, and the ability to share your library with others. Bruce D'Arcus has already suggested that they include RDFa to make that data consumable.
0So let's put some of this together. Say a professor makes a Zotero collection containing entries for the resources used in a class. Students can then also follow their professor in Zotero, letting the professor add additional interesting material via the collection (currently, I know some faculty member who do this, notable Steve Greenlaw, using delicious with somewhat brittle tags for each course). Probably that would work best with a tag like "recommended".
0Now, let's also imagine that the university library has already done with the University of Huddersfield did and released their circulation data, and it's been put into RDF (see Semantifying University of Huddersfield Library's Circulation Data). Using ISBN numbers, we should be able to link up the Zotero data with the circulation data.
0Keep going? Why not? Let's also link it up with data from OpenLibrary. OpenLibrary is working to create a URI for each book, and to do so in a FRBRized way. That allows the various editions or publications of a work to be tied together. So, for example, the Work page for Murder on the Orient Express shows many of the editions of it. For good measure, let's also bring in data from the RDF Book Mashup.
0Finally, after my post about semantifying Huddersfield library data, Johan Bollen at Los Alamos National Laboratory emailed me to point out that there is, in fact, an ontology that might work well for modeling circulation data: MESUR (MEtrics from Scholarly Usage of Resources). The project there has at least 10 BILLION triples about citations and usage. Yeah. Let's play.
0So what might be available? We've got a wonderful focal point of course resources with the Zotero collections. Importantly, this provides a common link for everyone in the class to use in any blog posts about the resource. (See Student Publishing and the Web Context and âWhereâs the Links?â: Pedagogy and the Blue Underline on my other blog, Semantic UMW, for why this is important).
0 That let's links (black arcs in the diagram below) from individual people, course groups, and library circulation data come together around items and libraries in Zotero. Following the model of the Huddersfield Library Data, that also associates the circulation data with courses and institutions. and if we can connect to MESUR, that will be about 2000 institutions.
0(Click for larger image -- opens in new window)
0Now, for the real stuff. Let's say my favorite book is "Murder on the Orient Express". From the OpenLibrary work page I can get a list of editions, then check in open Zotero libraries to see who else has it in their library (green arcs). And, for libraries that are associated with a course group, that leads back to the courses and institutions where it's studied (red arcs). Since the idea of a Giant EduGraph is also have info about the course groups up there (teacher, topics studied, course blogs, etc.), that gives a way to 1) find interesting courses studying the book and 2) trace back to blogs by people who are studying it.
0Say you find a blog you really like that way. Then go back from the blog to the blogger, then to that person's Zotero library to find more books you might like. It's not so much an (automated) recommendation system as it is a process, not unlike how we currently discover blogs by following links.
0The reality check is that there's quite a few details to fill in here. There's exposing all the Zotero info in some flavor of RDF, and the really big one is aggregating all of this data together so it can easily be browsed/queried (oh yeah, and there's the matter of a really good semantic web browser). But those will come along if people see this kind of system as something worthwhile.


Comments
Post new comment