Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/alexeyev/awesome-kyrgyz-nlp

Kyrgyz language processing software, models and datasets.
https://github.com/alexeyev/awesome-kyrgyz-nlp

List: awesome-kyrgyz-nlp

awesome-list corpus kyrgyz morphology natural-language-processing turkic turkic-languages

Last synced: 4 days ago
JSON representation

Kyrgyz language processing software, models and datasets.

Awesome Lists containing this project

README 

        
# Awesome Kyrgyz NLP [![Awesome](https://cdn.rawgit.com/sindresorhus/awesome/d7305f38d29fed78fa85652e3a63e154dd8e8829/media/badge.svg)](https://github.com/sindresorhus/awesome)



A curated list of awesome Kyrgyz language processing software, models and datasets. Inspired by [awesome-ML](https://github.com/josephmisiti/awesome-machine-learning). 



The main focus is on **open source** tools, **downloadable** data and research **papers with code**.



If you want to contribute to this list (please do), send me a pull request.

Also, a listed repository should be tagged as deprecated if:



* Repository's owners explicitly say that "this library is not maintained".

* Not committed to for a long time (2~3 years).



## Table of Contents



- [Awesome Kyrgyz NLP](#awesome-kyrgyz-nlp)

  - [Table of Contents](#table-of-contents)

  - [Datasets](#data)

  - [Pretrained models](#pretrained-models)

  - [Methods/Software](#software)

      - [Morphology](#morphology-s)

  - [Online Demos](#demos)

  - [Miscellaneous](#misc)





## Datasets



#### Corpora



* [Manas-UdS](http://fedora.clarin-d.uni-saarland.de/kyrgyz/index.html): 1.2M words, 84 literary texts, 5 genres: novel, novelette, epic, minor epic, and fairy tale; lemmata, PoS tags, rich per-text metadata.

* [kkWaC](https://www.sketchengine.eu/kywac-kyrgyz-corpus/): Kyrgyz corpus from the web, 19M words, Jan 2012

* [Kyrgyz in Leipzig Corpora Collecion](https://wortschatz.uni-leipzig.de/en/download/Kirghiz): Community data / Newscrawl (1M sentences) / Wikipedia (300K sentences)

* [TilCorpusu](https://tilcorpusu.org/): Kyrgyz corpus, 100M words, news+fiction, made public in July 2023



#### Character recognition

* [Kyrgyz language hand-written letters (Kyrgyz MNIST)](https://github.com/Akyl-AI/kyrgyz_MNIST): hand-written Kyrgyz alphabet letters collection for machine learning applications; original images (a total of 80213) have been transformed to 50x50 images, then to CSV format



#### Raw text

* [kloop corpus](https://github.com/kyrgyz-nlp/kloop-corpus): 16'826 articles (sqlite3 DB file) + crawler code



#### Morphology & Syntax

* [UD project comments](https://universaldependencies.org/tr/) on difficulties in Turkish language processing, might bring light to the question why parsing Kyrgyz is hard as well

* [KTMU's UD Treebank, 781 sentences](https://github.com/UniversalDependencies/UD_Kyrgyz-KTMU)

* [Small UD Treebank: 145 sentences (incl. 20 Cairo sentences), and ~ 100 sentences suggested by UD Turkic Group; a part of UD Turkic Treebank](https://github.com/UniversalDependencies/UD_Kyrgyz-TueCL); also note that the translations to English, Azerbaijani and Turkish are available

* [Verbal paradigms for Kyrgyz (100 Kyrgyz verbs fully conjugated in all tenses)](https://github.com/unimorph/kir) by Aytnatova Alima, annotation for Unimorph by E. Chodroff



#### Named Entity Recognition

* [WikiANN](https://huggingface.co/datasets/wikiann) has a Kyrgyz language part

* [KyrgyzNER](https://github.com/Akyl-AI/KyrgyzNER): [not published yet]



#### Text Classification

* [Kyrgyz Multi-Label News Classification](https://github.com/alexeyev/kyrgyz-multi-label-topic-classification): [not published yet]



#### Word Similarity Data

* [Kyrgyz Word Embedding Evaluation](https://github.com/alexeyev/kyrgyz-embedding-evaluation): [not published yet]



#### Instructions

* [Machine-Translated Alpaca](https://github.com/Akyl-AI/kyrgyz-alpaca): [Stanford Alpaca](https://github.com/tatsu-lab/stanford_alpaca) instructions translated into Kyrgyz using ChatGPT and Google Translate



#### Machine-readable dictionaries

* [Country names table](https://github.com/kyrgyz-nlp/country_names/blob/main/country_names_in_kyrgyz.tsv): Kyrgyz-Russian-English

* Thesaurus [KyrSpell](http://kyrgyzorfo.narod.ru/) (however, unpacking it seems to break the license)

* [Tatu Ylonen's enwiktionary-based dictionary](https://kaikki.org/dictionary/Kyrgyz/index.html) (also please see [the derived En-Ky Anki deck](https://ankiweb.net/shared/info/518863963) for language learners)





## Pretrained models

* [Polyglot morfessor](https://github.com/aboSamoor/polyglot/blob/master/docs/MorphologicalAnalysis.rst) —  pretrained [morfessor](http://www.cis.hut.fi/cis/projects/morpho/) model, number 6

* [fastText](https://fasttext.cc/docs/en/crawl-vectors.html) — 300-dimensional fastText vectors provided by the authors: [bin](https://dl.fbaipublicfiles.com/fasttext/vectors-crawl/cc.ky.300.bin.gz), [txt](https://dl.fbaipublicfiles.com/fasttext/vectors-crawl/cc.ky.300.vec.gz).

* [compressed fastText](https://github.com/avidale/compress-fasttext) —  [fasttext-ky-mini](https://zenodo.org/record/4905385/files/fasttext-ky-mini?download=1) prepared by Liebl Bernhard in 2021.

* [BERT-based NER](https://huggingface.co/murat/kyrgyz_language_NER) —  `bert-base-multilingual-cased` fine-tuned on Wikiann for NER on Kyrgyz. The author warns that this model is not usable and is built just as a proof of concept. Will be updated later.

* [Manas-GPT](https://github.com/janarosmonaliev/Manas-GPT/)  —  Janar Osmonaliev's fun personal project: training nanoGPT on Sayakbai Karalaev's version of Epic of Manas





## Methods/Software



* [spaCy](https://github.com/explosion/spaCy/tree/master/spacy/lang/ky) basic support: tokenization, stopwords, `like_num`



#### Morphology 

* [Kyrgyz for Apertium](https://github.com/apertium/apertium-kir/): morphological analysis and generation, PoS-tagging; installation script: [install_apertium_kir.sh](/install_apertium_kir.sh). A [much, much easier way](https://github.com/apertium/apertium-python/): `import apertium; apertium.installer.install_module("kir")`.

* [DEPRECATED] [kymopl](https://github.com/dens-den/kymopl): Kyrgyz morphology in Prolog



## Hate Speech detection

* [Jupyter Notebook for hate speech detection](https://github.com/kyrgyz-nlp/hate-speech-examples)



#### Other

* [Tilchi](https://github.com/IncorexLLC/Tilchi) electronic Russian-Kyrgyz dictionary, open source desktop application

* [ӨҮҢизатор](https://github.com/kyrgyz-nlp/letter_replacer): a proof-of-concept letter replacement Telegram bot demo code, fixes incorrect usages of 'О','У', 'Н' => 'Ө', 'Ү','Ң'

* [Number-to-words conversion](https://github.com/AzamatSoodaev/number-to-words-kg) (JavaScript) by [@AzamatSooldaev](https://github.com/AzamatSoodaev)

* [Number-to-words conversion](https://github.com/timursaurus/sandar) (TypeScript) by [@timursaurus](https://github.com/timursaurus)

* [Telegram bot for Kyrgyz morphological analysis](https://github.com/sasha-kir/kyrgyz_parser) by [@sasha-kir](https://github.com/sasha-kir) based on [Apertium data for Kyrgyz](https://github.com/apertium/apertium-kir/)





## Online Demos

* [Cyrillic-to-Latin online converter](http://www.transliteration.kpr.eu/ky/) based on [this resource](https://www.qyrgyz.com/post/kakoy-budet-kyrgyzskaya-latinitsa).





## Miscellaneous

* Kyrgyz NLP bibliography: [kyrgyznlp.github.io](https://kyrgyznlp.github.io)

* [Turkic Interlingua](https://github.com/turkic-interlingua) community and [SIGTURK](https://github.com/sigturk/) (ACL Turkic languages special interest group)

* A useful [Apertium](https://wiki.apertium.org/wiki/Kyrgyz)'s list of tools and other resources

* Online dictionaries and other useful resources on [el-sozduk.kg](https://el-sozduk.kg/)

* [Turkic languages-related resources](http://ddi.itu.edu.tr/en/toolsandresources) compiled by Dr. Gülşen Eryiğit and her team at Istanbul Technical University 

* [Data prepared by CSLT](http://index.cslt.org/mediawiki/index.php/ASR-nsfc-data): 128h speech, 163 speakers (100m/63f), transcription of the speech audio, lexicon in the word level; [link](https://pan.baidu.com/s/18jxLKo4YRWH5K3GWPWmGXQ) (requires extra steps, quote: `You should ask for license before you can download the datasets. Please send Email to [email protected] or [email protected] to get the license.`)



## Contributions to this list



[@golden-ratio](https://github.com/golden-ratio)