Topic on User talk:Pasleim

Could PLbot perhaps do a rerun of added descriptions that are erroneously upper case, like "Painter" instead of "painter" here in 2015? I've fixed a lot of them manually yet, but it's quite a boring task … Thanks in advance!

This description was imported from the Persondata template on English Wikipedia such that afterwards the template could be deleted.

I can correct descriptions, but would need either a list of words which should be written in lower case or a list of words which should be written in upper case.

Now I understand the origin of those descriptions.

1 SELECT DISTINCT ?item ?itemLabel ?description WHERE {
2   ?item (wdt:P106/wdt:P279*) wd:Q3391743.
3   ?item schema:description ?description.
4   SERVICE wikibase:label { bd:serviceParam wikibase:language "en". }
5   FILTER(STRSTARTS(?description, "Painter"))
6   FILTER((LANG(?description)) = "en")
7 }

(has to be copy pasted, I don't get Flow to render Template:SPARQL or a simple link properly) "Painter" alone concerns ~1700 cases given by WDQS when limited to occupation (P106):class of visual artist (Q3391743) items (times out otherwise; LIMIT doesn't help). In PLbot's contributions from 2015 I found an example for "Clergyman", probably there are many for "Actor", "Writer", "Basketball player" etc. The absence of the descriptions in the edit summaries and the timing out of WDQS handicap the compilation of such a list. Collecting words which should be written in upper case (mainly demonyms probably) sounds difficult and will have to many false positives I think.

Anyway, only fixing those ~1700 cases would help a lot yet. I guess it's better to have a list in the beginning than running it more times after more words have been found?

I corrected the descriptions with these terms.

For querying descriptions, using https://quarry.wmflabs.org/ is a more efficient way.

Thank you! And good to know that about Quarry!