June 16, 2008
Uncovering a Little More Digital Worlds Structure
Tweaking the Wordpress blog structure parser to get a GraphViz output was trivial, so here are a couple of snapshots for the reverses pingback graph (that is, if post B pingbacked post A, then I plot A as emergent linking forward to B).
So now I now have a couple of ways of looking at this stuff, what should I be looking for that may or may not be interesting...?;-)
Tags: digitalworlds, structure, blog structure
Posted by ajh59 at June 16, 2008 04:51 PMIt is interesting (and impressive), but I'm not sure it tells you what you need to know - how do users interact with the posts to assemble their own course?
Posted by: AJ Cann at June 17, 2008 09:32 AM"but I'm not sure it tells you what you need to know - how do users interact with the posts to assemble their own course?"
That's part of what I'm trying to work towards... (If you know all the answers, how come you aren't sharing them?! I don't even know what questions I'm trying to ask!; In fact, if you can help shape some tiny steps answerable questions, that would be really handy ;-)
So here's where I think I'm going at the mo... This "work" sort of complements the course analytics stuff. What I'm trying to do is muddle towards an understanding of a 'course graph' and how various resources are connected together.
Using web analytics, we can track how many people click through various links, what fragments of the graph they traverse in a study session, and so on.
In our online Relevant Knowledge short courses, the structure of the courses are pretty linear - pages end with a 'click here to go to the next page' link, and so on.
In the uncourse blog, by automatically annotating pages with trackbacks, we allow multiple forward navigational paths to emerge. I suspect (but don't know for sure) that at the moment: a) not many people click through on links contained within the body of a post; b) not many click through the trackback links. However, by styling the trackback links in a different way (e..g. to know more about x click here, for y, click there etc) "on-clicking" might be encouraged?
Over time, and with an active population, it would become possible to track user flow over connected posts (and potentially discover a consensus 'emergent course' from a set of resources.
Stephen Downes commented about by 'ad revenue' post the other day with a comment that "I conclude ... that we'll abandon the course production model." [ http://www.downes.ca/cgi-bin/page.cgi?post=44975 ]
So what I'm trying to do is a) [one extreme] with course analytics understand how people currently engage with linearly authored online materials; and b) [other extreme] try to understand how we can author in a natural an emergent way non-linear 'courses'.
Then there are all the steps that need adding in-between and round the edges: e.g. how can we help learners find structured/scaffolded paths through non-linearly authored, loosely coupled, emergently cross-referenced resources? Is there scope for instructors putting together anything other than linearly ordered courses (problem based learning and other resource based learning approaches try to get a handle on this).
My gut feeling is that we can in principle track use of online resources in a comprehensive way, and come to an understanding of 'course resource graph' structures that in some sense work for learners. But I don't know how to do it. So I'm trying lots of little things out..
I fear that I do know all the answers, I just don't know the questions. Sometimes I wake up in the middle of the night worrying about it ;-)
Posted by: AJ Cann at June 17, 2008 12:44 PM