Wikidata talk:Lexicographical data/Archive/2018/12

From Wikidata
Jump to navigation Jump to search
This page is an archive. Please do not modify it. Use the current page, even to continue an old discussion.

Improving the display of language of Glosses

Hello all,

Based on your feedback, the software is now displaying the languages of Glosses with the full name of the language, in the user's interface language, in reading mode. However, in editing mode, the field is still requesting a language code without suggesting the names of the languages. We're working on it :)

If you're interested, you can have a look at the tickets: phab:T203457, phab:T209931. Cheers, Lea Lacroix (WMDE) (talk) 14:22, 4 December 2018 (UTC)

Access the number of Forms and Senses for Lexemes

Hello,

If you're running tools, you can now access the number of Forms and Senses for Lexemes, through the page properties API and Special:PagesWithProp (choose wbl-forms or wbl-senses).

See the related ticket: phab:T199611.

As for now, the Lexeme having the biggest number of forms is dostatni (L21348) (84 Forms) and the one with the biggest number of Senses is żeński (L17561) (8 Senses). Some good job has been done on Polish words :)

Cheers, Lea Lacroix (WMDE) (talk) 12:55, 4 December 2018 (UTC)

@Lea Lacroix (WMDE): Thanks, but there is some bug in measurement. Lexeme mądry (L24242) has more forms (140) and вода (L189) has more senses (18). KaMan (talk) 13:06, 4 December 2018 (UTC)
Thanks for letting me know. Lea Lacroix (WMDE) (talk) 13:26, 4 December 2018 (UTC)
Update: in order to appear in the list, Lexemes need to have a new edit done on the page. That's why some of them don't appear yet. Lea Lacroix (WMDE) (talk) 17:17, 11 December 2018 (UTC)

Complexity of verb conjugations

I tried to create a WDQ query for verbs and their conjugations, but that is very complex.

There are multiple grammatical entries for

Should these entries be merged?

Furthermore present participle (Q24133704) is a redirect to present participle (Q10345583), but still used in many lexeme forms. I think all those entries need to be corrected.

The query that I made so far on b:SPARQL/WIKIDATA_Lexicographical_data is very complex, also because the conjugations are to be derived in different ways, for instance 1) first-person singular (Q51929218) or 2) first person (Q21714344) with or without singular (Q110786).

HenkvD (talk) 13:36, 9 December 2018 (UTC)

@HenkvD: I had thought I saw PLbot fixing some of those present participle (Q24133704) entries, as it does for redirects in statements on regular items, but that seems to have stopped? Or maybe it was just an experiment? Maybe we need a bot request for this... ArthurPSmith (talk) 14:01, 9 December 2018 (UTC)
I found a previous discussion in the archive of this talk page Wikidata_talk:Lexicographical_data/Archive/2018/07#Q13923816,_Q24133704,_and_Q24577575. Here you also mentioned the need to fix this by bot. Apparently this was not done automatically, HenkvD (talk) 17:37, 9 December 2018 (UTC)
@HenkvD: Note I have requested this to be done by bot here. ArthurPSmith (talk) 19:29, 11 December 2018 (UTC)

How to tell how many senses has a lexeme

Hi,

I just created glas (L40129) and glas (L40130) which is a colour of it's own in breton (the sea, the leaves of the trees, the sky are all "glas"). I'm wondering, should I create:

  • one sense for blue (Q1088) and green (Q3133) (and others colours) together
  • two senses fro blue (Q1088) and green (Q3133) (and others colours) appart
  • both? (that's strange but that's the option I'm leaning too right now, to correspond to the diversity of point of view in books and dictionaries)

For the context, you can read the articles linked to distinguishing blue from green in language (Q455950) (it's the same question for all Celtic languages - it's the same "glas" in Glasgow (Q4093) -, and some others languages like Russian in some order too).

Cheers, VIGNERON (talk) 18:09, 17 December 2018 (UTC)

Do people using the word distinguish those as two meanings (without using qualifiers such as in English we do with "light blue", etc.) or treat as one uniform "color"? I would lean toward just a single sense (encompassing all the colors lumped under that heading) unless its clear it's being used with distinct meanings in different contexts. ArthurPSmith (talk) 18:43, 17 December 2018 (UTC)
@ArthurPSmith: that's a good question. I would say for native speakers it's only one colour (and there is definitely no qualifier to distinguish the color of the trees or of the sky, here some attestations: « Gwez glas warne ’vel eur vantel » like a mantle of green trees on them and « En oabl glas an heol a lintre » The sun shines in the blue sky), but if you look into dictionaries, they often treat it as several colours.
And if I choose to use one sense only, should I add blue (Q1088) and green (Q3133) for item for this sense (P5137) or create a specific item about this color in Celtic languages?
Cheers, VIGNERON (talk) 19:26, 17 December 2018 (UTC)
@VIGNERON: Either create a new item, or propose the property "sense combines items".--Micru (talk) 23:39, 17 December 2018 (UTC)
The English term for this colour seems to be grue. That could have an item. —Rua (mew) 12:02, 18 December 2018 (UTC)
There's already new riddle of induction (Q1469453), but I don't know if that's appropriate for the colour. The Wikipedia links suggest that this refers to something else and could do with relabelling. —Rua (mew) 12:05, 18 December 2018 (UTC)
Not in Russian though (we have different words, at least different stems). May be cyan (Q180778) or aqua (Q372733) would fit? --Infovarius (talk) 16:44, 26 December 2018 (UTC)

Lingua Libre now adds audio pronunciation on Forms

Hello all,

First of all, if you don't know Lingua Libre: it's a very cool website to record the pronunciation of words, upload them on Commons and add them to Wiktionary or Wikidata.

Lingua Libre Bot was already adding links to the audio files on items when needed, and is now able to add the links to Forms as well!

To record words and add them to Lexemes, two steps are necessary:

  • generate lists of words from a query (documentation here)
  • record the words on Lingua Libre, the bot will take care of adding the links on Forms

Please note that these features are still beta. If you notice any issue, please contact User:0x010C.

Thanks to everyone who worked on this during last weekend's hackathon :) Lea Lacroix (WMDE) (talk) 11:37, 20 December 2018 (UTC)

Alphabetical order by the bot

Should´t the word "Évora" be alphabetically sorted before any word beginning with D? In another way, shouldn´t any word with a signal be treated, for alphabetically order, as without the signal? I notice a case were de Listeria Bot places the word "Évora" below "Viseu", for instance, in w:pt:Lista de Castelos de Portugal. Greetings and Merry Christmas! GualdimG (talk) 11:56, 24 December 2018 (UTC)

@GualdimG: it depends on the context, especially of the language. And anyway, your question doesn't seem to be related to Lexemes, you should ask on Wikidata:Project chat or directly to Listeria dev. Cdlt, VIGNERON (talk) 11:22, 29 December 2018 (UTC)
Ok, thank you, I had already ask at the talk of the bot. Greetings, GualdimG (talk) 11:29, 29 December 2018 (UTC)

Discussion on how to model adjectives

Please join the discussion here on how to model the relations between adjectives and associated nouns. ArthurPSmith (talk) 18:03, 27 December 2018 (UTC)

Requests for deletion

Hi,

Following the previous discussions about phonemes and graphemes, I created two requests for deletion :

A galon, VIGNERON (talk) 13:17, 29 December 2018 (UTC)

Retrieved from "https://www.wikidata.org/w/index.php?title=Wikidata_talk:Lexicographical_data/Archive/2018/12&oldid=838645552"