Apertium-kir

Kymorph is a morphological analyser/generator for the Kyrgyz language, currently working. It is intended to be compatible with transducers for other Turkic languages so that they can be translated between.

Installation

$ git clone https://github.com/apertium/apertium-kir.git

Adding analyses to conllu

This guide assumes that you

• have apertium-kir compiled and running,
• have a working recent version of python3, and
• are saving your files in the apertium-kir/corpus directory.

Here's what you need to do:

1. Get CG3 format of conllu data
   • From Annotatrix: click CG3 tab, click menu, download, rename file to .dep
2. Populate morphology in CG3 file from Apertium using ./corpora/add_morph.py script:
   • $ cat corpus.dep | ./corpora/add_morph.py > ./corpora/corpus_with_annotation.dep
   • Open ./corpora/corpus_with_annotation.dep and fix any errors
3. Create segmentation file with ./corpora/dep_to_seg.py script:
   • $ cat corpus_with_annotation.dep | ./corpora/dep_to_seg.py > ./corpora/corpus_with_annotation.seg
   • Open corpus_with_annotation.seg and fix any errors
4. Convert back to CoNLL-U format with ./scripts/conllise.py script:
   • $ python3 ./scripts/conllise.py -e apertium-kir.kir.udx ./corpora/corpus_with_annotation.dep ./corpora/corpus_with_annotation.seg

Current State

{{#set_param_default | corpus1 | None }} {{#set_param_default | corpus2 | None }} {{#set_param_default | corpus3 | None }} {{#set_param_default | corpus4 | None }} {{#set_param_default | corpus5 | None }} {{#set_param_default | corpus6 | None }} {{#set_param_default | corpus7 | None }} {{#set_param_default | corpus8 | None }} {{#set_param_default | corpus9 | None }} {{#set_param_default | corpus10 | None }}

• Number of stems: 14,424 {{#ifneq | | | () }}
• Disambiguation rules: 16
• Coverage: ~90.4%

{{#ifneq | azattyk2010 | None |

{{#ifneq | RFERL_corpora | | | }}

}}

{{#ifneq | azattyk2009 | None |

{{#ifneq | RFERL_corpora | | | }}

}}

{{#ifneq | bible | None |

{{#ifneq | | | | }}

}}

{{#ifneq | wp2011/04 | None |

{{#ifneq | | | | }}

}}

{{#ifneq | {{{corpus5}}} | None |

{{#ifneq | | | | }}

}}

{{#ifneq | {{{corpus6}}} | None |

{{#ifneq | | | | }}

}}

{{#ifneq | {{{corpus7}}} | None |

{{#ifneq | | | | }}

}}

{{#ifneq | {{{corpus8}}} | None |

{{#ifneq | | | | }}

}}

{{#ifneq | {{{corpus9}}} | None |

{{#ifneq | | | | }}

}}

{{#ifneq | {{{corpus10}}} | None |

{{#ifneq | | | | }}

}}

corpus	words	coverage
<nowinter>azattyk2010</nowinter>	azattyk2010	3.4M	~92.11%
<nowinter>azattyk2009</nowinter>	azattyk2009	4.1M	~92.04%
<nowinter>[[|bible]]</nowinter>	bible	174K	~92.25%
<nowinter>[[|wp2011/04]]</nowinter>	wp2011/04	545K	~85.37%
<nowinter>[[|{{{corpus5}}}]]</nowinter>	{{{corpus5}}}		~%
<nowinter>[[|{{{corpus6}}}]]</nowinter>	{{{corpus6}}}		~%
<nowinter>[[|{{{corpus7}}}]]</nowinter>	{{{corpus7}}}		~%
<nowinter>[[|{{{corpus8}}}]]</nowinter>	{{{corpus8}}}		~%
<nowinter>[[|{{{corpus9}}}]]</nowinter>	{{{corpus9}}}		~%
<nowinter>[[|{{{corpus10}}}]]</nowinter>	{{{corpus10}}}		~%