Comments on: EduGlu on Eduforge.org

By: Rob Wall

Rob Wall — Wed, 31 Dec 1969 17:00:00 +0000

Very cool, D’Arcy. This fits in nicely with some of the things I was thinking about when I was writing about building a school web site one blog at a time. My coding skills are a bit rusty, and were never much to begin with, but count me in.

By: Brian

Brian — Wed, 31 Dec 1969 17:00:00 +0000

My coding skills are non-existent, but I will happily cheerlead and discuss functionality.

And if the stuff Tyler and Enej is building turns out to be useful, we will be using Eduforge and whatever code that is useful will happily be shared.

By: D'Arcy

D'Arcy — Wed, 31 Dec 1969 17:00:00 +0000

Sweet. Wasn’t meaning to step on toes – I’m just using the Eduforge site to keep me motivated by something concrete. If aggRSSive 2006 is up to the task, I’m totally behind that. No need to build something else Just Because…

By: Brian

Brian — Wed, 31 Dec 1969 17:00:00 +0000

No stepping! No toes! This is awesome.

By: Scott Leslie

Scott Leslie — Wed, 31 Dec 1969 17:00:00 +0000

One fo the things that has really bugged me about the tagging business is lack of synonym support; the power of the approach lies in not dictating to people how they categorize things, but inevitably people end up using really different tags for the same thing (check out the plethora of ‘northern voices’ tags in flickr as an example). If we are serious about using this as a way to tie together disparate feeds for courses and other official education, then we need to work on this. I can see two ways to help; one is within the aggregator piece, a way for an instructor to indicate that various tags mean the same thing to them. Maybe just aggregating them together is enough. Another is to somehow replicate some of del.icio.us features on a more distributed scale. I really like the “recommended” tags and the “popular” tags feature of del.icio.us as a small step towards getting people to use ‘like’ tags, but how to do this across disparate blogging systems. Is there a way? Is it even worth it?

A small suggestion from the Genome sciences centre example was to annotate each post in a feed with a link that simply said “Interesting?” – this could help educate the aggregator and they were reporting very high usage of it after the initial deployment.

Something like http://taste.sourceforge.net/ (but I expect you are looking at non-Java solutions if you want anyone else to get into the development

Maybe talk with some of the existing open source CMS projects (Moodle, Atutor to be sure) to see what they have come up with to interface with institutional SIS – there may be usable pieces here to just incorporate into this.

By: D'Arcy

D'Arcy — Wed, 31 Dec 1969 17:00:00 +0000

I think we could be cool in providing a dynamically defined vocabulary of tags to be used. If a person tags outside of that set, then EduGlu wouldn’t/shouldn’t care about it.

So, you’d have to go to a personal page on EduGlu, and it would provide you with a list/cloud of tags being used in your class(es). You could then tag anything with one of those, and know it would automatically show up. Or, you could add a new tag to the cloud for use in a class.

I’m thinking of tags like “Chem355″ and “assignment” – things that should be rather static and concrete. The NV example is a good one, with “Northern Voice”, “northernvoice”, “northernvoice2006″ , “nv06″ and “moosecamp” all being used (along with other permutations).

Should we get into an arms race aimed at semantically mapping all known tags onto each other, or provide a set of tools to let groups define (and adhere to) a set of common and extensible tags?

By: Scott Leslie

Scott Leslie — Wed, 31 Dec 1969 17:00:00 +0000

Well, the latter I think, if at all. I definitely wasn’t endorsing any sort of large scale semantic mapping exercise or top-heavy approach, but there is an issue here to address. Check out http://www.louisrosenfeld.com/home/bloug_archive/000330.html for one of many recent discussions on this. The challenge becomes even larger if we truly do want to enable autonomy of choice of the authoring and reading tools people adopt, and don’t want to get into top-heavy standardization efforts. I don’t have the answer. I’m sure some smart person does. I’d ask Posthegemony Jon (sorry, forgot your last name) about his experiences using technorati to start to get some sense of the problems, and would definitely endorse talking with actual educators who are trying this already with various disparate toolsets to find out what the problems are that they are encountering.

By: D'Arcy

D'Arcy — Wed, 31 Dec 1969 17:00:00 +0000

maybe we should just use a LOM for each feed, with a taxonpath or something. Surely CanCore has already addressed this

One thing that might help is that we’d be using smaller subsets of the ’sphere in the aggregator. Where Technorati uses the entire known blogosphere as its source of data (and can therefore get a bit noisy) we’d be using only an infinitesimally small subset of feeds. Noise should be much smaller…

I like the concept of a metadata ecology though. Thanks for the link!

I totally agree about talking with Actual Educators, Actual Students, and perhaps even a token Actual Administrator or two to get perspectives on this from various angles…

By: Scott Leslie

Scott Leslie — Wed, 31 Dec 1969 17:00:00 +0000

Yeah, definitely the LOM. And RDF too. And definitely get the IEEE involved early on ;-#