CKAN
On this page, we describe how CKAN is used to catalog the datasets in the LOD Cloud. We start by showing how to list a dataset manually, then describe how some automation can assist the process. Next, we highlight some characteristics that CKAN collects using a manual process, to see if we can implement some [FAqT Services](FAqT Service) to help out.
This page has the following sections:
- Generic, Manual CKAN - clicking around on the website to manually create and describe a dataset.
- LOD Cloud specific metadata - understanding what extra metadata the lodcloud group requires (and, how to encode it in CKAN terminology).
- Reading from CKAN with curl - using HTTP to programmatically get the descriptions that you put in manually.
- Submitting a dataset with ckanclient - using a python client instead of curl directly.
- Submitting a dataset with RDF and SADI - wrapping the python client so that we can use RDF and avoid the CKAN-hack descriptions.
How to list a version of https://scm.escience.rpi.edu/svn/public/logd-csv2rdf4lod/data/source/contactingthecongress/directory-for-the-112th-congress/?
I. Go to http://thedatahub.org and create an account.
- You can create an account directly on CKAN, or use an Open ID, e.g. Google.
- After logging in, the Data Hub will show your API key, which can be used to programmatically retrieve and update CKAN data.
III. Review and accept the Open Database License
http://opendatacommons.org/licenses/odbl/1.0/
IV. Add a dataset
V. Have a name for your dataset
A description of the dataset should be provided, as well as the URL for your dataset on CKAN. The URL follows http://thedatahub.org/dataset/. Click "Add Dataset" as you done with the basic information.
(Note, the offer to "Upload or link some data now" only appears once, but it links to the "Edit : Resources" section of the dataset page, so you can get there yourself.)
VI. Provide some additional information
Then the page will be directed to add additional information. Additional information such as author, maintainer, contact info of the two as well as the link of the dataset should be provided. It is also preferred that data format (.ttl,.nt. or .rdf), namespaces (http://www.myData.org/), ontology (Dublin Core, FOAF), links to open dataset (number of links to DBpedia, freebase) and basic statistical information (total num of triples) are also given by the author to give a thorough view of your dataset.
V. Get the core dataset appeared online
There are 3 options to provide your data to the public.
- Link to URL
- Link to API
- Upload a file
The first option requests a link to your data that is already available on the web. The second option is to provide the link of your service (http://dbpedia.org/sparql). Finally, you may upload your data to CKAN so that they can take care of it.
When providing a Link to URL, the following metadata is requested:
The same metadata is requested when submitting a Link to API; though, most fields do not suit.
If you have a handful of similar datasets, it is a good idea to create a "CKAN group" for them. For example, the lodcloud CKAN group contains all datasets worth of being in the LOD Diagram.
Although "tagging" some CKAN datasets also groups them, creating a CKAN group is a little heavier weight. Each group gets its own page that lists its members.
- Go to http://thedatahub.org/group
- Click Create Group
Note that when you create a CKAN group, you must go into the administration settings to permit others to modify it. Groups are not "public edit" by default like dataset listings are.
- Metadata
Author, contact Info, description of the content of the data provide data consumer a basic idea of what the dataset is about.
- Basic Data Quality Preview
Number of links to the open dataset, ontology give the data consumer a basic confidence in using the data.
For details about what metadata they require and how it's encoded as both CKAN attributes and proper RDF vocabularies, see CKAN lodcloud RDF vocabulary.
The following figure illustrates the major resources that can be used to figure out how to publish a Linked Data set to the lodcloud group. Clicking on the image will lead to a higher resolution PDF version with links.
Going through the effort to list a dataset in CKAN has several advantages.
- CKAN provides a framework to maintain and publish information about your dataset, translating to less work for you.
- Your dataset can be discovered by those looking at CKAN
- You can add and modify your dataset listing yourself using a web browser
- You can add and modify your dataset listing programmatically (using REST!)
- Your RDF dataset can be listed in the lodcloud group, and thus become part of The Diagram
So far, this page has described how to list your dataset manually. Now we'll turn to how to access that information programmatically. Then, we'll look at submitting new datasets programmatically using the CKAN API, which is described at http://wiki.ckan.org/API with a tutorial at http://wiki.ckan.org/Using_the_API.
We'll use the Farmers Markets dataset as an example, since we listed a lot of its metadata and it is listed in lodcloud. First, we'll start with reading and writing at the CKAN api level, then move on to how CKAN has been interpreted as RDF by Will Waites and Keith Alexander.
Q: The attributes of a dataset listing?
Dataset's web page has the "package id" in its URL: http://thedatahub.org/dataset/farmers-markets-geographic-data-united-states
Change the base to get to the API: http://thedatahub.org/api/rest/package/farmers-markets-geographic-data-united-states
{
"maintainer":"",
"maintainer_email":"",
"id":"91d2c0de-75a4-4bb6-b260-bc2946e1be8b",
"metadata_created":"2007-04-10T21:19:38",
"relationships":[
],
"metadata_modified":"2011-12-05T11:35:55.419895",
"author":"Tim Lebo",
"author_email":"lebot@rpi.edu",
"download_url":"http://logd.tw.rpi.edu/source/data-gov/file/4383/version/2011-Nov-29/conversion/data-gov-4383-2011-Nov-29.ttl",
"state":"active",
"version":"2011-Nov-29",
"license_id":"cc-by",
"resources":[
{
"mimetype":"text/turtle",
"resource_group_id":"3bb8f6c7-25dc-4533-b4e5-0c84a984638e",
"hash":"",
"description":"Farmers Markets Geographic Data (United States) Version 2011-Nov-29, Enhancement 1",
"format":"text/turtle",
"url":"http://logd.tw.rpi.edu/source/data-gov/file/4383/version/2011-Nov-29/conversion/data-gov-4383-2011-Nov-29.ttl",
"cache_url":null,
"webstore_url":null,
"cache_last_updated":null,
"package_id":"91d2c0de-75a4-4bb6-b260-bc2946e1be8b",
"mimetype_inner":"text/turtle",
"webstore_last_updated":null,
"last_modified":"2011-11-29T11:40:00",
"position":0,
"size":"5772173",
"id":"99fce83a-f0c2-4efa-8421-59eac9fc4531",
"resource_type":"file",
"name":"Farmers Markets Geographic Data (United States) Version 2011-Nov-29, Enhancement 1"
},
{
"mimetype":"",
"resource_group_id":"3bb8f6c7-25dc-4533-b4e5-0c84a984638e",
"hash":"",
"description":"Rensselaer LOGD SPARQL Endpoint",
"format":"api/sparql",
"url":"http://logd.tw.rpi.edu/sparql",
"cache_url":null,
"webstore_url":null,
"cache_last_updated":null,
"package_id":"91d2c0de-75a4-4bb6-b260-bc2946e1be8b",
"mimetype_inner":"",
"webstore_last_updated":null,
"last_modified":null,
"position":1,
"size":null,
"id":"c71c409f-2da1-4ec4-983c-63494963fdb6",
"resource_type":"api",
"name":"Rensselaer LOGD SPARQL Endpoint"
},
{
"mimetype":"text/turtle",
"resource_group_id":"3bb8f6c7-25dc-4533-b4e5-0c84a984638e",
"hash":"",
"description":"voiD description",
"format":"meta/void",
"url":"http://logd.tw.rpi.edu/source/data-gov/file/4383/version/2011-Nov-29/conversion/data-gov-4383-2011-Nov-29.void.ttl",
"cache_url":null,
"webstore_url":null,
"cache_last_updated":null,
"package_id":"91d2c0de-75a4-4bb6-b260-bc2946e1be8b",
"mimetype_inner":"text/turtle",
"webstore_last_updated":null,
"last_modified":"2011-11-29T00:00:00",
"position":2,
"size":"183655",
"id":"8b29ea8f-c971-4c0a-a4bf-ae91d68f31bf",
"resource_type":"file",
"name":"voiD description"
},
{
"mimetype":"text/turtle",
"resource_group_id":"3bb8f6c7-25dc-4533-b4e5-0c84a984638e",
"hash":"",
"description":"RDF Schema",
"format":"meta/rdf-schema",
"url":"http://logd.tw.rpi.edu/source/data-gov/file/4383/version/2011-Nov-29/conversion/data-gov-4383-2011-Nov-29.void.ttl",
"cache_url":null,
"webstore_url":null,
"cache_last_updated":null,
"package_id":"91d2c0de-75a4-4bb6-b260-bc2946e1be8b",
"mimetype_inner":"text/turtle",
"webstore_last_updated":null,
"last_modified":"2011-11-29T00:00:00",
"position":3,
"size":"183655",
"id":"86d4c0d1-2418-4bae-80b3-d9bccd3337e4",
"resource_type":"file",
"name":"RDF Schema"
},
{
"mimetype":"text/turtle",
"resource_group_id":"3bb8f6c7-25dc-4533-b4e5-0c84a984638e",
"hash":"",
"description":"Turtle example link",
"format":"example/turtle",
"url":"http://logd.tw.rpi.edu/source/data-gov/file/4383/version/2011-Nov-29/conversion/data-gov-4383-2011-Nov-29.e1.sample.ttl",
"cache_url":null,
"webstore_url":null,
"cache_last_updated":null,
"package_id":"91d2c0de-75a4-4bb6-b260-bc2946e1be8b",
"mimetype_inner":"text/turtle",
"webstore_last_updated":null,
"last_modified":"2011-11-29T00:00:00",
"position":4,
"size":"183655",
"id":"c655d560-e174-474c-bd18-3a36ff9869e0",
"resource_type":"file",
"name":"Turtle example link"
}
],
"tags":[
"format-con",
"format-conversion",
"format-dc",
"format-ov",
"format-owl",
"format-void",
"format-wgs",
"geographic",
"lod",
"no-deref-vocab",
"provenance-metadata",
"published-by-third-party",
"vocab-mappings"
],
"groups":[
"lodcloud",
"datafaqs"
],
"name":"farmers-markets-geographic-data-united-states",
"license":"OKD Compliant::Creative Commons Attribution",
"notes_rendered":"<p>Longitude and latitude, state, address, name, and zip code of Farmers Markets in the United States, converted to RDF format.\n</p>\n<p>References <a href=\"http://logd.tw.rpi.edu/sparql.php?query-option=text&query=PREFIX+owl%3A+%3Chttp%3A%2F%2Fwww.w3.org%2F2002%2F07%2Fowl%23%3E%0D%0APREFIX+conversion%3A+%3Chttp%3A%2F%2Fpurl.org%2Ftwc%2Fvocab%2Fconversion%2F%3E%0D%0ASELECT+distinct+%3Fexternal%0D%0AWHERE+%7B%0D%0A++GRAPH+%3Chttp%3A%2F%2Flogd.tw.rpi.edu%2Fsource%2Fdata-gov%2Fdataset%2F4383%2Fversion%2F2011-Nov-29%3E++%7B%0D%0A++++%3Flocal+owl%3AsameAs+%3Fexternal%0D%0A++%7D%0D%0A%7D%0D%0Aorder+by+%3Fexternal&service-uri=&output=html&callback=&tqx=&tp=\">154 URIs</a> in DBPedia, GovTrack, and Geonames.\n</p>\n<p><a href=\"http://logd.tw.rpi.edu/sparql.php?query-option=text&query=PREFIX+conversion%3A+%3Chttp%3A%2F%2Fpurl.org%2Ftwc%2Fvocab%2Fconversion%2F%3E%0D%0APREFIX+ds4383_vocab%3A+%3Chttp%3A%2F%2Flogd.tw.rpi.edu%2Fsource%2Fdata-gov%2Fdataset%2F4383%2Fvocab%2F%3E%0D%0ASELECT+distinct+%3Fmarket%0D%0AWHERE+%7B%0D%0A++GRAPH+%3Chttp%3A%2F%2Flogd.tw.rpi.edu%2Fsource%2Fdata-gov%2Fdataset%2F4383%2Fversion%2F2011-Nov-29%3E+%7B%0D%0A++++%3Fmarket+a+ds4383_vocab%3AFarmersMarket+%0D%0A++%7D%0D%0A%7Dorder+by+%3Fmarket&service-uri=&output=html&callback=&tqx=&tp=\">7,223 farmers market URIs</a> dereference to RDF/XML (e.g., see <a href=\"http://validator.linkeddata.org/vapour?vocabUri=http%3A%2F%2Flogd.tw.rpi.edu%2Fsource%2Fdata-gov%2Fdataset%2F4383%2Fversion%2F2011-Nov-29%2FfarmersMarket_1019&classUri=http%3A%2F%2F&propertyUri=http%3A%2F%2F&instanceUri=http%3A%2F%2F&defaultResponse=dontmind&userAgent=vapour.sourceforge.net\">vapour report</a> for <a href=\"http://logd.tw.rpi.edu/source/data-gov/dataset/4383/version/2011-Nov-29/farmersMarket_1019\">farmersMarket_1019</a>).\n</p>",
"url":"http://explore.data.gov/Agriculture/Farmers-Markets-Geographic-Data/wfna-38ey",
"ckan_url":"http://thedatahub.org/dataset/farmers-markets-geographic-data-united-states",
"notes":"Longitude and latitude, state, address, name, and zip code of Farmers Markets in the United States, converted to RDF format.\r\n\r\nReferences [154 URIs](http://logd.tw.rpi.edu/sparql.php?query-option=text&query=PREFIX+owl%3A+%3Chttp%3A%2F%2Fwww.w3.org%2F2002%2F07%2Fowl%23%3E%0D%0APREFIX+conversion%3A+%3Chttp%3A%2F%2Fpurl.org%2Ftwc%2Fvocab%2Fconversion%2F%3E%0D%0ASELECT+distinct+%3Fexternal%0D%0AWHERE+%7B%0D%0A++GRAPH+%3Chttp%3A%2F%2Flogd.tw.rpi.edu%2Fsource%2Fdata-gov%2Fdataset%2F4383%2Fversion%2F2011-Nov-29%3E++%7B%0D%0A++++%3Flocal+owl%3AsameAs+%3Fexternal%0D%0A++%7D%0D%0A%7D%0D%0Aorder+by+%3Fexternal&service-uri=&output=html&callback=&tqx=&tp=) in DBPedia, GovTrack, and Geonames.\r\n\r\n[7,223 farmers market URIs](http://logd.tw.rpi.edu/sparql.php?query-option=text&query=PREFIX+conversion%3A+%3Chttp%3A%2F%2Fpurl.org%2Ftwc%2Fvocab%2Fconversion%2F%3E%0D%0APREFIX+ds4383_vocab%3A+%3Chttp%3A%2F%2Flogd.tw.rpi.edu%2Fsource%2Fdata-gov%2Fdataset%2F4383%2Fvocab%2F%3E%0D%0ASELECT+distinct+%3Fmarket%0D%0AWHERE+%7B%0D%0A++GRAPH+%3Chttp%3A%2F%2Flogd.tw.rpi.edu%2Fsource%2Fdata-gov%2Fdataset%2F4383%2Fversion%2F2011-Nov-29%3E+%7B%0D%0A++++%3Fmarket+a+ds4383_vocab%3AFarmersMarket+%0D%0A++%7D%0D%0A%7Dorder+by+%3Fmarket&service-uri=&output=html&callback=&tqx=&tp=) dereference to RDF/XML (e.g., see [vapour report](http://validator.linkeddata.org/vapour?vocabUri=http%3A%2F%2Flogd.tw.rpi.edu%2Fsource%2Fdata-gov%2Fdataset%2F4383%2Fversion%2F2011-Nov-29%2FfarmersMarket_1019&classUri=http%3A%2F%2F&propertyUri=http%3A%2F%2F&instanceUri=http%3A%2F%2F&defaultResponse=dontmind&userAgent=vapour.sourceforge.net) for [farmersMarket_1019](http://logd.tw.rpi.edu/source/data-gov/dataset/4383/version/2011-Nov-29/farmersMarket_1019)).",
"title":"Farmers Markets Geographic Data (United States)",
"ratings_average":null,
"extras":{
"links:govtrack":"52",
"triples":"130005",
"namespace":"http://logd.tw.rpi.edu/source/data-gov/dataset/4383/version/2011-Nov-29/",
"links:geonames-semantic-web":"50",
"shortname":"Farmers Markets (US)",
"links:dbpedia":"52",
"sparql_graph_name":"http://logd.tw.rpi.edu/source/data-gov/dataset/4383/version/2011-Nov-29"
},
"ratings_count":0,
"revision_id":"9c2545b8-ebf1-42bf-ac92-633ff5ceba3f"
}
Q: List all datasets in a group?
Group's web page has its id in its URL: http://thedatahub.org/group/datafaqs
Change the base to get to the API: http://thedatahub.org/api/rest/group/datafaqs
returns
{
"display_name":"DataFAQs",
"description":"",
"title":"DataFAQs",
"created":"2011-11-29T20:49:42.445597",
"state":"active",
"extras":{
},
"revision_id":"d9d3eca9-1a06-4d14-b269-5db44951a427",
"packages":[
"congresspeople",
"farmers-markets-geographic-data-united-states",
"white-house-visitor-access-records"
],
"id":"3c5f8753-6f42-45ca-a2be-9dd53eb679b8",
"name":"datafaqs"
}
Q: Add a tag to a dataset?
For example, tag http://thedatahub.org/dataset/white-house-visitor-access-records with "no-deref-vocab"
(the steps here work when changing "package" to "dataset" and using request header X-CKAN-API-Key instead of Authorization).
To modify a dataset listing, you need an API key, which is listed in your user home page after you log in to CKAN. For POST operations, the API key must be the value of an Authorization request header.
The example from the CKAN tutorial:
$ curl http://test.ckan.net/api/rest/package/test -d '{"name":"test", "title":"Changed Test package"}' -H "Authorization:your-api-key"
When you update an object, fields that you don’t supply will remain as they were before. (Model Formats) (though, this conflicts with recommendation via email to send EVERYTHING you get back)
We need to write the entire record, not a specific attribute. So we need to GET the whole thing, tweak it, then POST it back. Let's fall down to the command line.
curl -O http://thedatahub.org/api/rest/package/white-house-visitor-access-records
{ "maintainer":"", "maintainer_email":"", "id":"e59f5df0-05b4-40e5-99b1-f801c15e6e93", "metadata_created":"2007-04-10T21:19:38", "relationships":[ ],
"metadata_modified":"2011-11-29T14:02:44.873726",
"author":"",
"author_email":"",
"state":"active",
"version":"",
"license_id":"",
"resources":[
],
"tags":[
],
...
tweak it to say:
"tags":[
"no-deref-vocab"
],
...
Send it back up :
$ curl http://thedatahub.org/api/rest/package/white-house-visitor-access-records -d @white-house-visitor-access-records -H "Authorization:Aa8...Bb2"
"Access denied" :-(
also denied:
curl http://thedatahub.org/api/rest/package/white-house-visitor-access-records --data-urlencode @white-house-visitor-access-records -H "Authorization:Aa8...Bb2"
http://test.ckan.net/dataset/9f825e18-0e0c-40f6-8895-b5392c1ac18a
Now that we've covered how to use the core CKAN REST API to access and modify the metadata that we added manually, we can see how the python API can be used to perform the same operations. Although there are CKAN APIs for PHP, Javascript, Python, Perl and command-line, we focus here on Python because we are headed to making SADI services in Python. Each language-specific API provides a wrapper to the foundational REST API calls.
As an example, we'll tag e59f5df0-05b4-40e5-99b1-f801c15e6e93 / white-house-visitor-access-records with 'government' using CKAN's Python api.
First, we need to install the CKAN Python API:
sudo easy_install http://pypi.python.org/packages/source/c/ckanclient/ckanclient-0.9.tar.gz#md5=cb6d09eb2e60a01bce60c82c6c3a0c85
We submitted https://github.com/okfn/ckanclient/issues/15 regarding ~/.ckanclientrc not providing the api-key.
A SADI service to accept RDF-encoded tags and VoID, and submits it to CKAN: add-metadata.py. X_CKAN_API_Key is described at DATAFAQS environment variables.
export X-CKAN-API-Key="YOUR API KEY"
python add-metadata.py
class TagCKANDataset(sadi.Service):
...
name = 'add-metadata'
...
# Instantiate the CKAN client.
# http://docs.python.org/library/configparser.html (could use this technique)
key = os.environ['X_CKAN_API_Key']
if len(key) <= 1:
print 'ERROR: https://github.com/timrdf/DataFAQs/wiki/Missing-CKAN-API-Key'
sys.exit(1)
self.ckan = ckanclient.CkanClient(api_key=key)
Service will be available at: http://localhost:9090/add-metadata
Dumb down Turtle to RDF/XML to avoid bad python turtle parsers, and HTTP POST the RDF to the service:
$ rapper -g -o rdfxml sample-inputs/congresspeople-tagged-government.ttl > b.rdf
$ curl -H "Content-Type: application/rdf+xml" -d @b.rdf http://localhost:9090/add-metadata
Returns:
@prefix dcterms: <http://purl.org/dc/terms/> .
@prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#> .
<http://thedatahub.org/dataset/congresspeople> a <http://purl.org/twc/vocab/datafaqs#ModifiedCKANDataset>;
dcterms:modified "2011-12-16T17:19:26.635183";
rdfs:seeAlso <http://thedatahub.org/dataset/f4c2a8bb-6580-4919-98aa-617feb766b6c> .
Useful regex b/c dcterms:identifier are uuids AND pretty names (for same thing):
def is_id(self, id_string):
'''Tells the client if the string looks like an id or not'''
return bool(re.match('^[0-9a-f]{8}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{12}$', id_string))
I'd like to extend sadi.py to ckan-sadi.py that includes the ckan object and has a constructor:
def __init__(self):
sadi.Service.__init__(self)
# Instantiate the CKAN client.
# http://docs.python.org/library/configparser.html (could use this technique)
key = os.environ['X_CKAN_API_Key']
if len(key) <= 1:
print 'ERROR: https://github.com/timrdf/DataFAQs/wiki/Missing-CKAN-API-Key'
sys.exit(1)
self.ckan = ckanclient.CkanClient(api_key=key)
Are these services all in a common directory? You can include that code in the same directory as your services (ckan-sadi.py or ckan.py) and do:
from ckan-sadi import *
Since it'll be in your python path. You can have ckan-sadi.py import sadi.py and extend that class as you show below, then have your classes extend that one.
(Apr 2013)
CKAN lists dataset changes (e.g. klappstuhlclub)
http://docs.ckan.org/en/latest/api.html#get-able-api-functions describes how to construct the URL for the functions listed at http://docs.ckan.org/en/latest/ckan.logic.action.get.html
-
http://docs.ckan.org/en/latest/ckan.logic.action.get.html#ckan.logic.action.get.package_revision_list), and another one to get the revision details
- e.g. for klappstuhlclub
{
"help": "Return a dataset (package)'s revisions as a list of dictionaries.\n\n :param id: the id or name of the dataset\n :type id: string\n\n ",
"success": true,
"result": [
{
"id": "3d1b86ca-557a-4be2-82e6-2eeb2d9f0e4b",
"timestamp": "2011-10-05T15:32:01.037585",
"message": "",
"author": "139.18.255.65",
"approved_timestamp": null
},
{
"id": "998ebea5-d7e1-4c13-a76a-d7e9b8dd9688",
"timestamp": "2011-09-12T07:32:17.038294",
"message": "",
"author": "http:\/\/kurzum.myopenid.com\/",
"approved_timestamp": null
},
{
"id": "82b23f1a-5fdb-40f3-88a9-b4bbb1b46e76",
"timestamp": "2011-09-09T08:23:54.817967",
"message": "",
"author": "anjeve",
"approved_timestamp": null
},
{
"id": "20f06b51-2db0-49ca-888d-7facfc4d2a42",
"timestamp": "2011-09-09T08:18:29.275626",
"message": "",
"author": "anjeve",
"approved_timestamp": null
},
{
"id": "275cb117-0a25-4a66-bdf1-092bb99ea853",
"timestamp": "2011-09-09T07:56:31.842035",
"message": "",
"author": "anjeve",
"approved_timestamp": null
},
{
"id": "9e81f0ab-d0b5-49e1-a82c-11c5951a550f",
"timestamp": "2011-09-09T07:56:01.188565",
"message": "",
"author": "anjeve",
"approved_timestamp": null
},
{
"id": "28abe894-2cf1-4eb6-bd93-b14e0292042b",
"timestamp": "2011-09-09T07:53:14.447832",
"message": "creation",
"author": "http:\/\/kurzum.myopenid.com\/",
"approved_timestamp": null
}
]
}
Still need to find the call for e.g. http://datahub.io/revision/diff/farmers-markets-geographic-data-united-states?diff_entity=package&oldid=6278d190-33b2-4bf9-9e77-8a0f35d2540c&diff=9c2545b8-ebf1-42bf-ac92-633ff5ceba3f
- https://github.com/okfn/ckan/blob/master/ckan/lib/accept.py handles conneg to CKAN URIs.
- http://docs.ckan.org/en/latest/linked-data-and-rdf.html
- http://lab.linkeddata.deri.ie/2011/grefine-ckan/
- http://semantic.ckan.net/sparql is still alive as of Feb 2013, as a second-life maintained by OpenLink.
- https://github.com/srkunze/ckanext-semantic searches for the oldest VoID stats or oldest dataset without any VoID stats and then generated the corresponding VoID and save it to a given triplestore.
- Sean put together this overview of the CKAN code: https://gist.github.com/seanh/3414107
- https://github.com/open-data/ckanapi
- CKAN lodcloud RDF vocabulary for more discussion about the metadata that the lodcloud group requires and how it's encoded both as CKAN attributes and proper RDF vocabularies.
- Some DataFAQs Core Services use the CKAN API or the semantic.ckan form to get dataset listings and descriptions.
- FAqT Service if you're ready to write an evaluation service.
- SADI Semantic Web Services framework
- https://github.com/jimmccusker/twc-healthdata/wiki/Accessing-CKAN-listings discusses getting dataset listings from CKAN using semantic web principles (or lack thereof)