On 8/14/19 6:07 AM, Jérémie Roquet wrote:
Hi!
Le mer. 14 août 2019 à 01:10, Kingsley Idehen kidehen@openlinksw.com a écrit :
We have loaded Wikidata into a Virtuoso instance accessible via SPARQL [1]. One benefit is helping to understand Wikidata using our Faceted Browsing Interface for Entity Relationship Types [2][3].
That's great news, thanks!
Hi Jérémie,
You are welcome! See responses below.
Feedback always welcome too :)
So, I've eagerly tried a very simple SPARQL query with a huge result set, the complete version of which¹ I've known for several years to timeout in both the official Blazegraph instance and a personal Blazegraph instance with supposedly all time limits removed:
PREFIX wd: http://www.wikidata.org/entity/ PREFIX wdt: http://www.wikidata.org/prop/direct/
SELECT ?person WHERE { ?person wdt:P31 wd:Q5 }
… and while the Virtuoso instance manages to answer pretty quickly, it seems that it's cutting the result set at 100k triples. Is it the expected behavior?
Yes.
If so, I suggest you show that in the UI because apart from the improbable round number of triples, it's not obvious that the result set is incomplete (in this case, the LDF endpoint tells us that there should be around 5,4M triples²).
Thanks again!
¹ ie. using the wikibase:label service ² https://query.wikidata.org/bigdata/ldf?subject=&predicate=wdt%3AP31&...
If you open up your browser's inspector you will see:
cache-control: max-age=3600 content-encoding: gzip content-type: text/html; charset=UTF-8 date: Wed, 14 Aug 2019 16:47:47 GMT expires: Wed, 14 Aug 2019 17:47:47 GMT server: Virtuoso/08.03.3315 (Linux) x86_64-generic-linux-glibc25 VDB status: 200 strict-transport-security: max-age=15768000 vary: Accept-Encoding x-sparql-default-graph: http://www.wikidata.org/ *x-sparql-maxrows:* 100000
In addition, note that Virtuoso has an "Anytime Query" feature [1][2] that it uses to drive a "Fair Use" policy that ensures an endpoint is able to handle a cocktail of query types from users and bots. This is also how we handle DBpedia and DBpedia-Live instances [3]. Naturally, HTTP response metadata will also inform you when this kicks in.
[1] http://docs.openlinksw.com/virtuoso/anytimequeries/
[2] http://vos.openlinksw.com/owiki/wiki/VOS/VirtTipsAndTricksAnytimeSPARQLQuery
[3] https://wiki.dbpedia.org/public-sparql-endpoint