I have been doing it from open street map data using a python script.

At first it generates about 100gb of zipped ttl files, imports them with a specialized blazegraph tool and afterwards uses direct sparql statements to do batch updates when new data becomes available. 

https://github.com/Sophox/sophox/tree/main/osm2rdf


On Wed, Apr 14, 2021, 5:07 PM Henry Rosales <hrosmendez@gmail.com> wrote:
Hi all!

I hope you are doing very well! I’m working on constructing an RDF dataset, which I’m ingesting to Wikibase and leverage its ecosystem: visualization, SPARQL endpoint, a script for dump creation, etc. For ingestion, I’m using the API (explained in  https://www.wikidata.org/w/api.php), and it is going well, except for the time performance.

I was exploring the Quick Statements module of Wikibase, which allows the ingestion through the UI. However, I need a way to ingest data automatically. I wonder if Wikibase/Wikidata has any script for bulk loads.

Any comment or suggestion will be welcome.

Best Regards,
Henry Rosales

_______________________________________________
Wikitech-l mailing list
Wikitech-l@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikitech-l