Talk:Abstract Wikipedia/Community:2023-04-21

@Dušan Kreheľ: as regards some of your directly raised questions (though I have no clue what "NGS" is supposed to mean, nor can I tell what compelled you to write the content of the 'Questions' section at the bottom):

• What is it like to have an extensive dictionary? Have you taken a look at what Wikidata lexemes can contain?.
  • D.: If it's so good, why isn't there e.g. any dictionary that uses that data (rhetorical question)? This would also be suitable for practical demonstrations of the use of these data (i.e. leave the theoretical plane only).
• At what language level? At every language level. :-)
• The user will have to create everything. What is the ratio of investment versus result? The investment put into certain types of lexemes has yielded interesting results thus far.
• Will a database of some established phrases be created somewhere? Not "established phrases", but certainly idiomatic ones and fixed phrases separately defined can have lexemes, and equivalents to these phrases might be transformed into different languages as needed.
• Simply at all, can the text be a mix of automatic versus manual? There is intended to be a visual distinction made between automatically generated text and manually typed text; as for a mixture of the two, that's less clear, although I'd prefer for as little mixing as possible (so that improvements to the text generation system are directed at lexemes and functions rather than at raw text).
• What if I want to set somewhere that the word ma should be the link in the generated sentence? As I am an sk-0, you're going to need to be much more specific about this.
  • D.: I updated the document now, so u look now.

Mahir256 (talk) 19:36, 21 April 2023 (UTC),
D.: --Dušan Kreheľ (talk) 08:46, 22 April 2023 (UTC)