All statements of a specific datatype: monolingual text (not important for querying says Lydia)
Description
Description
| Status | Subtype | Assigned | Task | ||
|---|---|---|---|---|---|
| Resolved | • AKhatun_WMF | T282790 [EPIC] Get estimates for dropping data from Wikidata in case of Blazegraph catastrophic failure | |||
| Open | None | T288264 Get estimates for all Wikidata statements of a specific datatype |
Event Timeline
Comment Actions
@Lydia_Pintscher
Is this ticket asking for counts of various datatype used in WIkidata? Both URI and literals.
Does wikitech:User:AKhatun/Wikidata_Basic_Analysis#Object help?
Comment Actions
@AKhatun_WMF: Basically Wikidata's Properties have a datatype. The possible ones are listed on https://www.wikidata.org/wiki/Special:ListDatatypes. What would be interesting to know is how many statements we have for each datatype so that we can understand how much we'd gain if we consider removing all statements of a particular datatype.
Examples:
- https://www.wikidata.org/wiki/Q2#P31 <- 3 statements for datatype "Item"
- https://www.wikidata.org/wiki/Q2#P7471 <- 1 statement for datatype "External Identifier"
I am not seeing that in the analysis you linked but maybe I am overlooking something.
Comment Actions
Basically Wikidata's Properties have a datatype.
Ah, datatype of properties.
I am not seeing that in the analysis you linked but maybe I am overlooking something.
The one I listed is for datatype of objects, so you didn't miss anything.
Thank you for clarifying! It should be fairly easy to find out as well :)
Comment Actions
It seems it's not that easy. The queries for popular datatypes (including Monolingualtext) time out, see https://w.wiki/4GED. It works for unpopular types like TabularData though: https://w.wiki/4GEG.