More on Blog@Case Topics
Okay, I've talked about this Blog@Case Topics thing before; and I am going to talk about it again.
I have this grand vision of collaboration. Where people who talk on the topic of Law can go to http://blog.case.edu/topics/law and find others on the Blog@Case system talking about the same (or similar stuff). 47 people all talking about law, all posting under the category Law, and all discovering and collaborating with one another via keeping an eye on http://blog.case.edu/topics/law, which presents itself as both a human-viewable HTML page and a syndicated news feed that people could subscribe to in their aggregator of choice or in the portal. This, obviously, extrapolates itself out to:
http://blog.case.edu/topics/eces340http://blog.case.edu/topics/itshttp://blog.case.edu/topics/physicshttp://blog.case.edu/topics/compscihttp://blog.case.edu/topics/usghttp://blog.case.edu/topics/whatever
Everbody posting and collaborating and discovering one another and having a great time. So-called "Social Taxonomy" (or "Folksonomy") for the Blog@Case system just like del.icio.us and Flickr tags but for Blog@Case blog entries.
But, right now, there is not so much cross-pollination in the system. People post under categories, but there is not much overlap in what categories people are using — both in namespace and in semantics. You have the "General" (http://blog.case.edu/topics/general) and "work" (http://blog.case.edu/topics/work) categories that get some cross-blog-goodness, but as of right now, that is about it. You have other people posting under "Photos" while others post under "Photo's" or "Pics" or "Pictures."
The system may start self-organizing once it is noticed and people begin to use it, but I may need to perform some vicious data mining and language-stemming if it is to be of practical use. And, I hadn't really planned on doing that.
We'll see, I suppose.
Hi Jeremy, have you taken a look at K-Collector?
http://w4.evectors.it/
Cheers.
Jeremy,
Sounds like great minds think alike. We have wanted to do the same thing for a while. We've always had pages which show all entries for a particular category, but no way of really encouraging people to use matching categories. So, I built this page last week: http://blogs.warwick.ac.uk/blogbuilder/sysadmin/topCategories.spr
It will hopefully help people notice which categories to use and provide them an easy way to explore them (idea from Flickr).
what if you just focus on teaching people to use del.icio.us to keep track of posts by others in the Blog@Case community that were of interest to them, and to tag them with Blog@Case and whatever other categories they think are relevant?
Then, a year from now, you'd have five people who have started to tag things Blog@Case and Law routinely, (along with a lot of other tags, too), and Blog@Case and Management, and Blog@Case and Medicine, etc. Then you can use those socially generated categorization schedmes as RSS feeds or the basis for some kind of indexing by topic and date and author of post.
Jermey,
Not sure of how deep your control of this application is, but... Have you considered tweeking the interface for adding new catergories to a persons blog? Have a couple of example categories, a link to a list of popular categories and a link explaining why using the same categories as other people is cool. Maybe even do some cool autocompletion like Google Suggest.
Trying to automate categorizing could turn out to be pretty ugly so you could try some social engineering in the interface to get people to chose the categories you want them to.
Steve
Yes, I got as far as this page where I read:
I believe I, then, closed my browser.
Ahhh, that's nice stuff, Kieran.
That's an interesting idea, and that could very well happen on its own. If it did happen (i.e. people started tagging such in that way), I would certainly begin using it (in some nefarious way or another).
Complete control. It's just some Perl code, after all. I can muck with it all day and night and then some.
That could get pretty involved.
"Recently Popular" or "Popular" in general? Or, even, "Popular Categories You Have That Are Often Used by Others." Man, it could get really hairy.
That would work so long as you labor under the assumption that people follow helpful links to documentation that would benefit them. Unfortunately, I do not labor under that assumption because I, myself, don't follow such links.
But, how could I determine what categories to suggest to them? I could scan the content they are creating, see that they have an IMG tag in there and suggest "Photo." Or, I could parse out any hyperlinks they have included in the content, determine who else in the Blog@Case system has linked to those same URLs, and suggest those categories other people have used... The Law of Diminishing Returns is started to become enacted here.
We're pretty close to doing exactly what you're describing over at http://blog.educause.edu/
Going forward, I hope to introduce features like those mentioned in this comment.
http://blog.case.edu/jms18/2005/03/15/more_on_blogcase_topics#625
Very interesting stuff!
Thanks,
Matt
Getting your URL structured the right way can be a little difficult. Especially if there are competing interests as there maybe here..