For a while we've been having discussions about 'better data' - and
one of the issues has always been the intelligent tagging of activity
data with keywords.
In a dataset that may not have the activity of your dreams (search),
presenting credible options/alternatives is important, and keywords
can help us do that.
A while ago we found the
http://tagthe.net webservice that generates
tags from blocks of text (and URLs), and built a quick demo (see:
http://plings.info/wiki/index.php/Tag_Getter). Then we did nothing!
The problem with this was that there were lots of tags you would
normally want to reject - if only there were some sort of check list
of useful tags....
Well now there is. We reckon this page:
http://plings.info/wiki/index.php/Keywords,
is a pretty good place to start, and Ben has generated this code:
http://plings.info/wiki/index.php/Keyworderer
as a demo of how you could run your data through a matching process
and add keywords to your activities - auto-magically!
By using a pre-defined set of tags you can reject (or store and
examine later) tags without human intervention. Of course you can
always send tags that are not on our list - and we'd love to hear of
tags you think we're missing - but in the face of a good description
field with no tags - this might be preeety useful.
Check it out and let us know if this is of any use to you...
Cheers
David