Difference between revisions of "Apertium-kaz"

From Apertium
Jump to navigation Jump to search
(Undo revision 38277 by Sushain (Talk))
(Undo revision 38273 by Sushain (Talk))
Line 15: Line 15:
   
 
== Current State ==
 
== Current State ==
  +
* Number of stems: {{#lst:Apertium-kaz/stats|stems}}
{{LangStats | lang = kaz | corpus1 = Әуезов | corpus2 = bible | corpus3 = azattyq2010 | corpus4 = wp2011 | corpus5 = quran}}
 
  +
* Coverage: ~{{:Apertium-kaz/stats/average}}%
  +
  +
{| class="wikitable"
  +
|-
  +
! corpus !! words !! coverage
  +
|-
  +
|[[Әуезов corpus|Әуезов]]
  +
|align="right"| {{#lst:Apertium-kaz/stats|Әуезов-words}}
  +
| ~{{#lst:Apertium-kaz/stats|Әуезов-coverage}}%
  +
|-
  +
| bible
  +
|align="right"| {{#lst:Apertium-kaz/stats|bible-words}}
  +
| ~{{#lst:Apertium-kaz/stats|bible-coverage}}%
  +
|-
  +
| [[RFERL corpora|azattyq]] 2010
  +
|align="right"| {{#lst:Apertium-kaz/stats|azattyq2010-words}}
  +
| ~{{#lst:Apertium-kaz/stats|azattyq2010-coverage}}%
  +
|-
  +
|wp 2011-11
  +
|align="right"| {{#lst:Apertium-kaz/stats|wp2011-words}}
  +
| ~{{#lst:Apertium-kaz/stats|wp2011-coverage}}%
  +
|-
  +
| quran
  +
|align="right"| {{#lst:Apertium-kaz/stats|quran-words}}
  +
| ~{{#lst:Apertium-kaz/stats|quran-coverage}}%
  +
|-
  +
|}
   
 
== To-do ==
 
== To-do ==

Revision as of 01:36, 9 January 2013

Kazmorph is a morphological analyser/generator for Kazakh, currently under development. It is intended to be compatible with transducers for other Turkic languages so that they can be translated between.

Installation

kazmorph is currently located in incubator/apertium-kaz.

Dependency tree

  • hfst (svn ≥r1916)
    • foma
      • flex
  • apertium
    • lttoolbox

Current State

  • Number of stems: 36,595
  • Coverage: ~94.5%
corpus words coverage
Әуезов 155K ~92.89%
bible 577K ~95.29%
azattyq 2010 3.2M ~95.07%
wp 2011-11 850K ~90.72%
quran 107K ~96.71%

To-do

Improve coverage

  • Causitives
  • collective numbers
  • fix demonstratives
  • vowel harmony of single-syllable words with у and и

Future

  • run tests on morphophonology