Difference between revisions of "Uighur and Turkish/Work plan"

From Apertium
Jump to navigation Jump to search
Line 207: Line 207:
 
==Plan by Weeks==
 
==Plan by Weeks==
   
# 30% coverage
+
# 45% coverage, adding new stems to bidix and monodix
  +
# 65% coverage, adding new stems to bidix and monodix
# Basic CG
 
# 40% coverage
+
# 67% coverage, Basic CG
  +
# 70% coverage, Adding new morphemes to uig.lexc, writing twol rules for them
# Transfer
 
  +
# 75% coverage, Adding new morphemes to uig.lexc, writing twol rules for them
# 50% coverage
+
# 78% coverage, Transfer, CG
# Transfer, lexical selection, 65% coverage
 
# CG, 80% coverage
+
# 82% coverage, CG, lexsel
# Transfer, lexsel, 84% coverage
+
# 84% coverage, Transfer, lexsel
# Transfer
+
# 85% coverage, Transfer, lexsel
# CG, Transfer
+
# 85% coverage, CG, Transfer
# Transfer, lexsel, 86% coverage
+
# 86% coverage, Transfer, lexsel
# Transfer, 88% coverage
+
# 88% coverage, Transfer, CG
 
# Preparing text for annotation, evaluation
 
# Preparing text for annotation, evaluation
 
# Annotating the Uyghur corpus, %90 coverage
 
# Annotating the Uyghur corpus, %90 coverage

Revision as of 07:47, 9 May 2018

Week Cov. goal CG goal Transfer goal Lexsel goal Corpusvoc goal Done? Coverage Errors Checkpoint Comments
April 23-29 45% 5 200000 45.5 197742 Good work!
April 30 - May 6 65% 10 198000 ½ 65.6 197742 Good coverage, insufficient CG rules
May 7-13 67%
May 14-20 70%
May 21-27 75%
May 28-June 3 78%
June 4-10 82% 20
June 11-17 84% Eval 1
June 18-24 85%
June 25 - July 1 85% 20
July 2-8 86%
July 9-15 88% Eval 2
July 16-22 89%
July 23-29 90%
July 30 - August 5 91%
August 6-14 92% 0 Final Evals

Plan by Weeks

  1. 45% coverage, adding new stems to bidix and monodix
  2. 65% coverage, adding new stems to bidix and monodix
  3. 67% coverage, Basic CG
  4. 70% coverage, Adding new morphemes to uig.lexc, writing twol rules for them
  5. 75% coverage, Adding new morphemes to uig.lexc, writing twol rules for them
  6. 78% coverage, Transfer, CG
  7. 82% coverage, CG, lexsel
  8. 84% coverage, Transfer, lexsel
  9. 85% coverage, Transfer, lexsel
  10. 85% coverage, CG, Transfer
  11. 86% coverage, Transfer, lexsel
  12. 88% coverage, Transfer, CG
  13. Preparing text for annotation, evaluation
  14. Annotating the Uyghur corpus, %90 coverage
  15. Annotating the Uyghur corpus, %90 coverage, Writing paper
  16. Writing paper

Plan Outline

  • Post-application period:
    • Facilitating MT of a text from Uyghur to Turkish.
  • Community-bonding period:
    • bidix words, up to 50%
  • Month 1:
    • Writing scripts
    • Adding words to bidix, get coverage to around 80%
    • Chunking
    • Transfer rules
    • Begin CG for UIG
  • Month 2:
    • POS tagging/constraint grammar
    • Transfer rules
    • Get CG rules up to 100, ~50% disambiguation
    • >90% coverage
  • Month 3:
    • Creation of an Annotated Corpus