Thank you very much for all your answers everyone.

I'm also trying to have a solid understanding of the wikidata data model. How references are stored and structured, how revisions are stored, etc.

My goal is to download the historic dump from wikidata and parse it into tables to analyse the references and their edits through time.

Can anyone point me to a resource which explains it?

Best regards,

Gabriel Maia 
Data Scientist and Developer

gabrielmaiarocha@gmail.com
gabrielmaia7.github.io
+55 85 99430 5370

Sent from the tiny tiny keys of my mobile phone

On Mon, Jan 20, 2020, 19:08 Gabriel Altay <gabriel.altay@gmail.com> wrote:
Hi Gabriel, its Gabriel :) 

If the query service is timing out on you and you'd like to work with a raw Wikidata JSON dump, you might find this package I've been working on useful.  

https://qwikidata.readthedocs.io/en/stable/index.html

You could use it to pull out all of the reference/source statements. There is an example of accessing references in the "entity" section of the docs, 

 https://qwikidata.readthedocs.io/en/stable/entity.html

On Mon, Jan 20, 2020 at 12:52 PM James Heald <jpm.heald@gmail.com> wrote:
On 20/01/2020 14:19, Nicolas VIGNERON wrote:
> Le lun. 20 janv. 2020 à 14:14, Eugene Alvin Villar <seav80@gmail.com> a
> écrit :
>
>> On Mon, Jan 20, 2020, 9:06 PM Nicolas VIGNERON, <
>> vigneron.nicolas@gmail.com> wrote
>>
>>> In a nutshell, you can safely assume that only the property 'stated in'
>>> matters.
>>>
>>
>> I think 'inferred from' (P3452) should also be considered as an internal
>> reference.
>>
>
> Probably but this property is not often used.
>
> I managed to make a query work: https://w.wiki/FqH (number of use of the 76
> properties used in the references space for items located in Paris, the
> last part is to have a small sample and avoid timeout ; also, maybe the
> SPARQL code could be optimized, I just wanted to have an overview and
> confirm my assumption that only a few property are widely used)
>

Here's a version of the query with labels for the properties, making its
output a bit more readable:
     https://w.wiki/Fr9

Note that for external references, it's very common just to give the
identifier in the external database, via the appropriate property
(perhaps accompanied by a "retrieved" = <date>, but usually not.

   -- James.

_______________________________________________
Wikidata mailing list
Wikidata@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikidata
_______________________________________________
Wikidata mailing list
Wikidata@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikidata