COINS as RDF

[Ian Dickinson]

We have made a version of the COINS data dump that HMT released earlier
this year available as RDF, using the data cube vocabulary. The sparql
endpoint is available here:

http://openuplabs.tso.co.uk/sparql/gov-coins

and documentation here:

http://data.gov.uk/resources/coins

This should be regarded as a beta release, and any feedback, comments
questions or suggestions welcome.

One thing to note: the redirects have not yet been put in place, so the
URI's don't resolve yet. This will be fixed shortly.

Ian

--
____________________________________________________________
Ian Dickinson Epimorphics Ltd, Bristol, UK
mailto:i...@epimorphics.com http://www.epimorphics.com
cell: +44-7786-850536 landline: +44-1275-399069
------------------------------------------------------------
Epimorphics Ltd. is a limited company registered in England
(no. 7016688). Registered address: Court Lodge, 105 High St,
Portishead, Bristol BS20 6PT, UK

[Kingsley Idehen]

On 11/26/10 7:06 AM, Ian Dickinson wrote:

We have made a version of the COINS data dump that HMT released earlier this year available as RDF, using the data cube vocabulary. The sparql endpoint is available here:

http://openuplabs.tso.co.uk/sparql/gov-coins

and documentation here:

http://data.gov.uk/resources/coins

This should be regarded as a beta release, and any feedback, comments questions or suggestions welcome.

One thing to note: the redirects have not yet been put in place, so the URI's don't resolve yet. This will be fixed shortly.

Ian

Ian,

What's the URL for the RDF dump?

-- 

Regards,

Kingsley Idehen	      
President & CEO 
OpenLink Software     
Web: http://www.openlinksw.com
Weblog: http://www.openlinksw.com/blog/~kidehen
Twitter/Identi.ca: kidehen 

[Kingsley Idehen]

On 11/26/10 7:06 AM, Ian Dickinson wrote:

> We have made a version of the COINS data dump that HMT released
> earlier this year available as RDF, using the data cube vocabulary.
> The sparql endpoint is available here:
>
> http://openuplabs.tso.co.uk/sparql/gov-coins
>
> and documentation here:
>
> http://data.gov.uk/resources/coins
>
> This should be regarded as a beta release, and any feedback, comments
> questions or suggestions welcome.
>
> One thing to note: the redirects have not yet been put in place, so
> the URI's don't resolve yet. This will be fixed shortly.
>
> Ian
>

Ian,

Also, is there a URL for machine readable variants of data cube ontology?

--

Regards,

Kingsley Idehen
President& CEO

[Richard Cyganiak]

On 26 Nov 2010, at 15:36, Kingsley Idehen wrote:
> Also, is there a URL for machine readable variants of data cube
> ontology?

http://purl.org/linked-data/cube#

Best,
Richard

[Kingsley Idehen]

On 11/26/10 10:42 AM, Richard Cyganiak wrote:
>
> On 26 Nov 2010, at 15:36, Kingsley Idehen wrote:
>> Also, is there a URL for machine readable variants of data cube
>> ontology?
>
> http://purl.org/linked-data/cube#

Richard,

Any chance of some rdfs:isDefinedBy triples connecting the Ontology to
the Classes and Properties it describes? This would make the Ontology
more navigable.

Links:

1.
http://linkeddata.uriburner.com/describe/?url=http%3A%2F%2Fpurl.org%2Flinked-data%2Fcube%23ComponentSet
-- example of what I currently see; an isDefinedBy relation would offer
more holistic navigation

2.
http://linkeddata.uriburner.com/describe/?url=http%3A%2F%2Fpurl.org%2Fgoodrelations%2Fv1
-- example with GoodRelations ontology

3.
http://linkeddata.uriburner.com/describe/?url=http%3A%2F%2Fpurl.org%2Fgoodrelations%2Fv1&p=12002
-- example that hones into isDefinedBy since my instance has a lot of
GoodRelations instance data which affects location of specific page
showcasing isDefinedBy effects.

[Ian Dickinson]

Hi Dan,
> This is great, excuse the nitpic:
>
> "select ?deptLongName ?rcLabel (COUNT(?amount) as ?n) (SUM(?amount) as
> ?total) ..." ... in the docs looks like some version of SPARQL I'm not
> familiar with. Maybe SPARQL 1.1 drafts, or Virtuoso/Jena extensions? I
> thought in 1.0 we couldn't do fancy things in the SELECT clause.
Yes, sorry, this is SPARQL-1.1. I tested all the queries against Andy's
TDB, forgetting that not all triple stores support 1.1 yet. My bad. I've
taken an action to update the documentation.

[Ian Dickinson]

Hi Kingsley,

> Ian,
>
> What's the URL for the RDF dump?

You frame the question as though it's axiomatic that there will be an
RDF dump as well as the SPARQL end-point and (in-progress) JSON API. I'm
not sure that it is.

While I did the translation of COINS to RDF, TSO are handling the
hosting and I would say it's their call whether to provide the raw dump.
And I can see why they might hesitate - it's a 2.4Gb tar.gz file, or
130Gb of uncompressed n-quads. When I sent the development versions of
the data to the TSO team to mount into their store, it was easier to
drop the files on a dvd and post it than to upload over ADSL.

Still, no harm to ask - Tom, Paul, what do you think?

[Dan Brickley]

On Fri, Nov 26, 2010 at 5:35 PM, Ian Dickinson <i...@epimorphics.com> wrote:

>> What's the URL for the RDF dump?
>
> You frame the question as though it's axiomatic that there will be an RDF
> dump as well as the SPARQL end-point and (in-progress) JSON API. I'm not
> sure that it is.
>
> While I did the translation of COINS to RDF, TSO are handling the hosting
> and I would say it's their call whether to provide the raw dump. And I can
> see why they might hesitate - it's a 2.4Gb tar.gz file, or 130Gb of
> uncompressed n-quads. When I sent the development versions of the data to
> the TSO team to mount into their store, it was easier to drop the files on a
> dvd and post it than to upload over ADSL.

A little while back, a few of us via the public-lod list started
experimenting with the use of Bittorrent to share DBpedia data dumps.
That experiment fizzled, but perhaps it's worth revisiting the idea.
It would need a few orgs with bandwidth and storage to get things
rolling, but perhaps has some potential for sharing the burden of this
kind of data hosting?

Dan

[Kingsley Idehen]

On 11/26/10 11:35 AM, Ian Dickinson wrote:
> Hi Kingsley,
>> Ian,
>>
>> What's the URL for the RDF dump?
> You frame the question as though it's axiomatic that there will be an
> RDF dump as well as the SPARQL end-point and (in-progress) JSON API.
> I'm not sure that it is.
>
> While I did the translation of COINS to RDF, TSO are handling the
> hosting and I would say it's their call whether to provide the raw
> dump. And I can see why they might hesitate - it's a 2.4Gb tar.gz
> file, or 130Gb of uncompressed n-quads.

Yes, but SPARQL Endpoint != RDF Dump. Of course, it implies existence of
a SPARQL accessible data space on an HTTP network (e.g. Web), but that's
only part of the picture re. access options for Linked Data.

As you can see with DBpedia, there are dumps and SPARQL endpoints.

> When I sent the development versions of the data to the TSO team to
> mount into their store, it was easier to drop the files on a dvd and
> post it than to upload over ADSL.
>
> Still, no harm to ask - Tom, Paul, what do you think?

Please ask them what the options are re. dumps. I am sure CKAN is
capable of holding such. Ditto others players in the Data Mart space :-)

BTW - I am able to get at the data via SPARQL-FED, but this isn't the
norm at this point in time, so dumps are still helpful to others that
may want to create or enhance their own Linked Data spaces.
>
> Ian
>
>

--

Regards,

Kingsley Idehen
President& CEO

[Richard Cyganiak]

Kingsley,

On 26 Nov 2010, at 16:14, Kingsley Idehen wrote:
>> http://purl.org/linked-data/cube#
>
> Richard,
>
> Any chance of some rdfs:isDefinedBy triples connecting the Ontology
> to the Classes and Properties it describes? This would make the
> Ontology more navigable.

Thanks for the request! I added the triples.

> 1. http://linkeddata.uriburner.com/describe/?url=http%3A%2F%2Fpurl.org%2Flinked-data%2Fcube%23ComponentSet
> -- example of what I currently see; an isDefinedBy relation would
> offer more holistic navigation

It seems to be caching the old version?

Best,
Richard

[Kingsley Idehen]

On 11/27/10 2:24 PM, Richard Cyganiak wrote:
> Kingsley,
>
> On 26 Nov 2010, at 16:14, Kingsley Idehen wrote:
>>> http://purl.org/linked-data/cube#
>>
>> Richard,
>>
>> Any chance of some rdfs:isDefinedBy triples connecting the Ontology
>> to the Classes and Properties it describes? This would make the
>> Ontology more navigable.
>
> Thanks for the request! I added the triples.
>
>> 1.
>> http://linkeddata.uriburner.com/describe/?url=http%3A%2F%2Fpurl.org%2Flinked-data%2Fcube%23ComponentSet --
>> example of what I currently see; an isDefinedBy relation would offer
>> more holistic navigation
>
> It seems to be caching the old version?

Yes, here is a little cleanup trick:

Go to the SPARQL endpoint, then add the pragma: define get:soft
"replace", which tells the sponger to overwrite current cache. Example:

define get:soft "replace"
select distinct * from <http://purl.org/linked-data/cube> where {?s ?p ?o}

I've executed the above, so you're now set :-)

Kingsley