Folksonomies
2008-04-04 20:26:35.002031+00 by
ebradway
7 comments
In a recent thread, I mentioned the idea of creating a folksonomy for Flutterby. A quick Google search turned up very little in the way of easy widgets for building a folksonomy out of an existing database. Anyone have experience with creating folksonomies?
I'd be interested in seeing both folksonomies based on an "extrapolated" ontology of keywords (i.e., common words of significance in the database that are not like "the" or "in") and based on the ontology created by entry keywords (which aren't used uniformly on Flutterby).
The results, of course, would be these ontologies with tags scaled according to either a count in the back-end database or a query on Google (keyword site:flutterby.com) linking to the appropriate search result.
Thoughts?
[ related topics:
Databases
]
comments in ascending chronological order (reverse):
#Comment Re: made: 2008-04-04 20:48:46.985861+00 by:
dws
"Programming Collective Intelligence" (O'Reilly) has some clustering algorithms that might give you a start, at least for finding interesting terms.
#Comment Re: made: 2008-04-04 21:10:06.888856+00 by:
Dan Lyke
My issue with "folksonomies" is summed up in the observation that the Dewey Decimal System is a semester long course for librarians.
And that nobody but me uses the categorization features on Flutterby entries...
But don't pay attention to my negativity, there's probably good stuff to be had out there, the way Flickr users use tags seems pretty awesome, for instance.
#Comment Re: made: 2008-04-04 21:39:05.993716+00 by:
Dan Lyke
Quick addition: I'd *love* it if you'd want to apply some of these mining theories to better categorization of the Flutterby archives...
#Comment Re: made: 2008-04-05 00:44:50.147151+00 by:
vanderwal
You may want to look at FreeTag as a build in ( http://code.google.com/p/freetag/ ).
#Comment Re: made: 2008-04-07 17:13:29.25701+00 by:
ebradway
Part of my interest in this is looking at how a "generated" ontology (frequently used words -{common English words}) compares to an "intentional" (but poorly used) ontology (the categories in Flutterby).
And that you take issue with the Dewey Decimal System being a semester course for librarians tells me that you don't understand the significance or the difficulty in creating an ontology that maintains semantic relationships across epistemological boundaries.
The Flutterby archives, I think, are unique in that there are a some distinct epistemologies that would be evident in a "generated" ontology. Specifically, I would expect, if the ontology were partitioned into politics, social relationships and the natural world, to see lexicons of "Libertarianism", "Queer", and "Positivism".
#Comment Re: made: 2008-04-07 17:41:02.74193+00 by:
Dan Lyke
Oh, no, I don't take issue with the Dewey Decimal System being a semester long course, I use the difficulty of reasonably learning how to classify using it (or the LC, or whatever) as an example of why it's unlikely that a random collection of people are going to spontaneously generate a generally useful categorization system.
#Comment Re: made: 2008-04-09 19:05:09.00445+00 by:
ebradway
I'm not looking to categorize the archives - rather - I'm looking to categorize Flutterby!