Page MenuHomePhabricator

Warn users when trying to add redundant statements using instance-of/subclass-of in Wikidata
Open, Needs TriagePublic

Description

(1) It's considered that a statement <QA> P31 (instance of) <QB> is redundant with a statement <QA> P31 (instance of) <QC> if there's a statement <QC> P279* (subclass of, recursively) <QB>.

  • Example of SPARQL query that lists QAs for a particular case of (1) where QB = Q12280 (bridge): The Query

(2) It's considered that a statement <QA> P279 (subclass of) <QB> is redundant with a statement <QA> P279 (subclass of) <QC> if there's a statement <QC> P279* (subclass of, recursively) <QB>.

Note that (2) could also be applied to other properties like P131 (located in the administrative territorial entity).


These redundant statements are usually added by all kinds of users (sometimes, using automated tools and, sometimes, not using them) who aren't aware of being adding redundant statements or who aren't aware that redundant statements should be avoided.

https://www.wikidata.org/w/index.php?title=Special:Contributions/Abi%C3%A1n&offset=20161002091931&limit=20&target=Abi%C3%A1n

Event Timeline

abian created this task.Oct 2 2016, 10:33 AM
Restricted Application added a subscriber: Aklapper. · View Herald TranscriptOct 2 2016, 10:33 AM
Esc3300 added a subscriber: Esc3300.Oct 2 2016, 1:09 PM

somewhat odd that you'd revert the initial statement on an item stating that it's redundant. It might be that you are not aware how statements are added.

abian added a comment.EditedOct 2 2016, 1:58 PM

somewhat odd that you'd revert the initial statement on an item stating that it's redundant. It might be that you are not aware how statements are added.

I fear I'm not understanding what you exactly mean.

Anyway, for clarification, the idea is to warn the user that he/she is introducing a redundant statement so that he/she can make one of the following decisions:

  • simply doing nothing and not adding that new statement because the old one is more specific,
  • adding the new statement (which is more specific) and removing the old one (which is more general),
  • adding the new statement and fixing the statements using subclass-of in the items <QB> and <QC>, which have wrong values.