Kevin
See answers inline below. Hope I've understood the questions correctly.
By the way, I've now matched all the parties to their respective
wikipedia entries, so that is now returned in all versions (HTML, RDF, etc).
On 19/03/2010 02:08, Kevin Wells wrote:
In message<[email protected]>
CountCulture<[email protected]> wrote:
Thought members of this list might be interested in knowing that the
Open Election Data Project website is now live at
http://OpenElectionData.org. The project is a group effort (with support
>from many different bodies) to encourage and enable Local Authorities to
publish their election results as open data (there is no open, public
database of local election results, only a commercial one).
It's also worth pointing out that as part of this we needed to have
linked data resource URIs for all the political parties, and so we've
done that in the form: http://openelectiondata.org/id/parties/6 (for the
Labour Party, for example).
I have a couple of questions.
The above link seems to redirect to
<http://openelectiondata.org/parties/6> missing the id part of the url
will that have different information on it at a later date?
Will the parties pages have more information on them at a later date?
For example number of MPs, councillors etc.
The URI http://openelectiondata.org/id/parties/6 is essentially an ID so
that it is clear when they are talking about the Labour Party (whatever
words are use on the ballot paper, and there are many alternative
names), they are specifically talking about the same party.
You could just leave it at that, and have the URI as a theoretical URI
that represents an entity or resource, that didn't resolve or return
anything, and to a degree it would work, being an id that all councils
(or anyone else) could use, that would allow clearer identification than
just a number or string (because it includes its own namespacing).
However, the standard practice in linked data is to dereference the
Resource URI and return information about the resource. This is usually
done by 303 redirects to a document about the resource, with the format
of the document depending on what you're asking for (in the Accepts
header). If you ask for RDF/XML in the Accepts header, you'll get
information in RDF/XML, similarly for XML and HTML (the default). You
can try this out:
curl -H "Accept: application/xml" http://openelectiondata.org/id/parties/6
curl -H "Accept: application/rdf+xml"
http://openelectiondata.org/id/parties/6
curl -H "Accept: text/html" http://openelectiondata.org/id/parties/6
All of these return basic information about the party -- the name,
alternative names, the Electoral Commission page about the party, and
either the wikipedia page, or the dbpedia one, depending on whether
you're getting HTML or RDF/XML, and in the case of the dbpedia one we
say the resource is the sameAs the one at the dbpedia URI.
We'll also be adding information about the party's website in the
future, and maybe other core info. However, there's no intention at the
moment to do a list of councillors or MPs, as the purpose of the URI is
to allow people to use it to uniquely identify a party. You could
already use the dbpedia URIs, except for the fact that a large
proportion of the parties on the list don't have wikipedia entries (and
therefore don't have dbpedia entries).
What this does allow other sites to do (TWFY, for example) when they
start publishing linked data is to use these URIs to refer to the
parties, or use their own URIs and on the documents about those URIs say
that their URI for the Labour Party is the sameAs the
http://openelectiondata.org/id/parties/6.
OpenlyLocal will be publishing a list of local authority election
results published as a result of the project, will be doing just that.
So using something as sameAs.org or Sindice etc you will be able to go
from the resource URI http://openelectiondata.org/id/parties/6 to get
information from dbpedia, openlylocal, and anyone else using the URI.
Hope this helps. I suppose the short answer would have been, these URIs
were created primarily to allow LAs to publish their election results as
linked data, but having created them they can become much more powerful,
and have many other uses.
Chris
The final digit is the Electoral Commission id for the party, so it
makes it easy, given that, to work out the Electoral Commission page for
the party (or indeed other info they hold, such as emblems). There'll be
a very lightweight API added for the parties in the next few days, so
you can query it by name (or alternative name) and get back an XML
response. All open data of course.
I've had a quick play with the xml feed and have produced a list of all
the parties, with links to their respective page on your site, and I'm
thinking of linking them to the Electoral Commission site of their page
as well.
An interesting site, with lots to play with.
Cheers
Chris
--
-------------------------------------------------------
OpenlyLocal :: Making Local Government More Transparent
http://openlylocal.com
Blog: http://countculture.wordpress.com
Twitter: http://twitter.com/CountCulture
_______________________________________________
Mailing list [email protected]
Archive, settings, or unsubscribe:
https://secure.mysociety.org/admin/lists/mailman/listinfo/developers-public
