[Sigia-l] Eliminating categories in favour of tagging
Alexander Johannesen
alexander.johannesen at gmail.com
Fri Mar 10 06:19:24 EST 2006
On 3/10/06, andrew at friendlymanual.com
> It depends how you allow the end users to do the tagging
Hey, *everything* depends. :) But yes, if you try to move from a
generic plane to a technical one (implementation specific) you can
alter what tagging means for sure.
> - if it is free text,
> then quite frankly, you deserve what you get. It can be as controlled
> as you like if you provide them with a bunch of checkboxes
> populated with CV terms and make them submit non-preferreds for
> addition separately.
Sure thing. Many degrees of control. One thing I haven't talked too
much about, though, is the structural inference needed in tagging vs.
the structure inherit in categories and taxonomies, but each of these
(and other variants) have their pro's and con's.
For example, tagging as we know it through Flickr and del.icio.us (I
love the last site; it has come up with a wonderful and geek-friendly
way to teach me how to write "delicious"!) is mostly unstructured raw
tagging; the structure that del.icio.us delivers is inferred, that is,
they need to analyse how the tags play together to convert it from a
graph-structure to a tree-structure. In a taxonomy this is default
behaviour (and also hence faster to traverse and infer on in a
computer system), but you lose the facetted aspect of them (well, lose
them in a sense that again you need to analyse semantics between
structures to derive at a semantic map of how they fit together. It's
all very interesting, actually :).
Anyways, this is why I love Topic Maps; I can have both (and more) at
the same time, in the same map. Gotta love that.
Alex
--
"Ultimately, all things are known because you want to believe you know."
- Frank Herbert
__ http://shelter.nu/ __________________________________________________
More information about the Sigia-l
mailing list