[ckan-discuss] ckan.net stats page

Mon Feb 22 11:08:36 GMT 2010

I've added a page to ckan.net giving statistics about the data it holds:
http://ckan.net/stats

It should help us in looking how well the metadata has been collected
so far and have implications for the design of the site. Here are my
observations:

* Packages are being added at about 100 per month in the past 6
months, which is about double of what it was in the 2 years before
that.
* Jonathan leads the way in 'owning packages' - almost half those in
CKAN. Not many other people are logging in or taking ownership.
* The top tags seem to show key properties of the packages, rather
than the subjects they are about. Maybe we should try and separate
these two, or shift some tags into groups?
* I suspect package ratings are a bit scuppered by spam - we don't ask
people to login. Only about a dozen ratings total have been given by
people logged in, but hopefully that would rise if we forced people to
login to rate (like Amazon does).
* The most edited package is top because of suffering lots of spam in
Nov 2007. Should we purging those edits?
* Groups seem to get healthy use

David