[ckan-discuss] Release of DataCatalogs.org to map open data around the world

Rufus Pollock rufus.pollock at okfn.org
Sat Jul 2 12:08:43 BST 2011


On 1 July 2011 18:16, Philip Ashlock <phil at openplans.org> wrote:
> I'm really excited about datacatalogs.org and glad to see it looking so
> polished.
>
> I've been trying to maintain a list of data catalog aggregators (sites like
> datacatalogs.org) which could perhaps be used to build a checklist of the
> people we can get in touch with to help consolidate at datacatalogs.org or
> to set up a scrapper or a notification mechanism to see when they are
> updated.
>
> http://wiki.civiccommons.org/Initiatives#Open_Government_Data_Catalog_Aggregators
>
> One valid argument that I think many will present against exclusively
> consolidating at datacatalogs.org is the limited means by which listings are
> organized and displayed at datacatalogs.org. The lack of a map was
> mentioned, but I don't think that is nearly as important as providing a
> simple way of grouping and therefore browsing using an intuitive and
> consistent hierarchy or taxonomy rather than solely relying on search and
> adhoc tagging.

All of this is actually there but not turned on at the moment. There
is full faceting capability (building on underlying solr search) in
CKAN (see e.g. on http://ckan.net/ this page http://ckan.net/package)
and we've put all the geo structure in we need to do geo faceting. The
issue is (I believe -- as I wasn't directly involved in the core dev)
is that in the sprint to get it launched this week we've haven't yet
turned all of this on.

> Perhaps this can be done by trying to better enforce a consistent identifier
> for Spatial Coverage, either by using a predefined directory or API which
> provides such things or attempting to automatically geocode the input using
> free APIs, even perhaps hosting a GeoDict install:
> http://www.datasciencetoolkit.org/developerdocs#geodict

We're already doing it I believe :-) (thought not using geodict).

Rufus



More information about the ckan-discuss mailing list