search

apertium / apertium-kir

Apertium linguistic data for Kyrgyz
GNU General Public License v3.0
14 stars 3 forks source link

Entry point for non-linguists #11

Closed jumasheff closed 1 year ago

jumasheff commented 2 years ago

Hey @jonorthwash @ftyers !

Today I showed this repo to a person who is trying to build an app for Kyrgyz language learners. He is going to use Udahin's ky-ru, ru-ky dictionaries (you can view files here) and asked me to help him with POS tagging. I showed the README examples and also mentioned Apertium's UD building efforts. I told him that the best thing to do is to start helping Apertium with UD and learn how a proper tagging should look like. But I don't think a person without special training will be happy to see and use Apertium/UD tags.

Questions:

  1. Are there low-hanging fruits in this project?
  2. How can a regular person contribute to this project?
  3. Are there online courses that teach things related to UD? 🙏

I think that Apertium is a fundamental tool that paves a path for Kyrgyz language to the world of computational linguistics and eventually AI (in its broad sense). I'd even teach a course on CompLing using Apertium at our local universities (only after getting educated myself, of course).

jumasheff commented 2 years ago

UPD: Found this crash course on Linguistics: https://youtube.com/playlist?list=PL8dPuuaLjXtP5mp25nStsuDzk2blncJDW I don't expect that it'll help me to understand Apertium better, though. Just an intro stuff.

jonorthwash commented 1 year ago

@jumasheff, somehow I missed this!

@IlnarSelimcan put together a guide for Kazakh speakers to expand apertium-kaz. I believe part of it is here, but I remember there being more.

Since needs change in terms of where to focus attention within a given transducer, it's a little hard to put together a good guide. I'd be very happy to talk (or work) with someone on anything related to Kyrgyz transducer or UD, though!

jumasheff commented 1 year ago

@jonorthwash Hey, thank you!

I am going to create a products NER for Kyrgyz and I think it's a good chance to learn everything related to CoNLL-U. I'll get back to you later with actual sentences :)