Difference between revisions of "User:Firespeaker/GSoC2014/Workplan"

From Apertium
Jump to navigation Jump to search
 
(12 intermediate revisions by the same user not shown)
Line 5: Line 5:
 
* A '''production-ready''' release of '''tur-kir'''
  
* A '''production-ready''' release of '''tur-kir'''
 
** Translates tur→kir and kir→tur with consistently <20% WER
  
** Translates tur→kir and kir→tur with consistently <20% WER
−
** Trimmed coverage for tur and kir ≥90%
 +
** Trimmed coverage for tur and kir ≥85%
 
* A '''stable release''' of '''uzb-tur'''
  
* A '''stable release''' of '''uzb-tur'''
 
** Translates tur→uzb and uzb→tur with consistently <25% WER
  
** Translates tur→uzb and uzb→tur with consistently <25% WER
−
** Trimmed coverage for tur and kir ≥80%
 +
** Trimmed coverage for tur and uzb ≥80%
  +
* While '''bidix size''' is not built into the goals, the trimmed coverage numbers can be seen as a more relevant proxy for the same basic idea.
   
 
== Plan ==
  
== Plan ==
Line 30: Line 31:
 
** four 200-word texts in each kaz, kir, tur, uzb
  
** four 200-word texts in each kaz, kir, tur, uzb
 
** four 500-word texts in each kaz, kir, tur, uzb
  
** four 500-word texts in each kaz, kir, tur, uzb
  +
| {{Workeval5|3}}
  ⚫
|
  +
* Reworked apertium-tur verb morphology on paper
  +
* Much better disam in apertium-tur
  +
* Have a bunch of texts, not many post-edited
  +
|
  +
* Need to rework transfer rules for apertium-tur
 
|-
  
|-
 
!colspan="2" style="text-align: right"|community bonding period<br />21 April - 19 May
  
!colspan="2" style="text-align: right"|community bonding period<br />21 April - 19 May
Line 41: Line 49:
 
* one 200-word tur-uzb text to <10% WER
  
* one 200-word tur-uzb text to <10% WER
 
* one 200-word uzb-tur text to <10% WER
  
* one 200-word uzb-tur text to <10% WER
  +
|{{Workeval5|2}}
  +
|
  +
* Implemented reworking of apertium-tur verb morphology
  +
|
  +
* Around, but lost time due to end of semester and conferences
 
|-
  
|-
 
! 1 !! 19 - 24 May
  
! 1 !! 19 - 24 May
Line 47: Line 60:
 
* one 200-word kir-kaz text to <10% WER
  
* one 200-word kir-kaz text to <10% WER
 
* work on kir CG and lrx
  
* work on kir CG and lrx
  +
|rowspan="3"| {{Workeval5|3}}
  +
|
  +
* Fixed some apertium-tur phonology to go with morphology rework
  +
|
  +
* Mostly worked on LREC poster
 
|-
  
|-
 
! 2 !! 25 - 31 May
  
! 2 !! 25 - 31 May
Line 54: Line 72:
 
* work on kaz CG and lrx
  
* work on kaz CG and lrx
 
* work on tur CG and lrx
  
* work on tur CG and lrx
  +
|
  +
|
  +
* At LREC, worked a lot on apertium-uig and apertium-kaz-uig
 
|-
  
|-
 
! 3 !! 1 - 7 June
  
! 3 !! 1 - 7 June
Line 61: Line 82:
 
* work on uzb CG and lrx
  
* work on uzb CG and lrx
 
* work on tur CG and lrx
  
* work on tur CG and lrx
  +
|
  +
* Fixed a small handful of transfer issues in apertium-tur-kir
  +
|
  +
* Got apertium-uig in a state for others to work with
  +
* Completed eval for last few weeks
 
|-
  
|-
 
! 4 !! 8 - 14 June
  
! 4 !! 8 - 14 June
Line 68: Line 94:
 
* work on kir CG and lrx
  
* work on kir CG and lrx
 
* start testvoc nouns for all pairs
  
* start testvoc nouns for all pairs
  +
|{{Workeval5|4}}
  +
|
  +
* Brought coverage of apertium-tur-kir on SETimes corpus up by 2%
  +
* Brought testvoc of apertium-tur-kir on SETimes corpus down from 10.39% to 0.22%
  +
|
 
|-
  
|-
 
! 5 !! 15 - 21 June
  
! 5 !! 15 - 21 June
Line 76: Line 107:
 
* work on tur CG and lrx
  
* work on tur CG and lrx
 
* continue testvoc nouns for all pairs
  
* continue testvoc nouns for all pairs
  +
|{{Workeval5|3}}
  +
|
  +
* Brought testvoc of apertium-tur-kir on SETimes corpus down from 0.22% to 0.04%
  +
|
  +
* Made and presented poster for Morphology Fest
 
|-
  
|-
 
! 6 !! 22 - 28 June
  
! 6 !! 22 - 28 June
Line 84: Line 120:
 
* work on uzb CG and lrx
  
* work on uzb CG and lrx
 
* continue testvoc nouns for all pairs
  
* continue testvoc nouns for all pairs
  +
|{{Workeval5|0}}
⚫
|-
  
  +
|
⚫
! 7 !! 29 June - 5 July
  
 
|
  
|
  +
* Moving and getting situated week
−
* finish testvoc nouns for all pairs
  
  +
* (break for personal reasons)
 
|-
  
|-
−
!colspan="2" style="text-align: right"|midterm eval<br />July 6
 +
!colspan="2" style="text-align: right"|midterm eval<br />29 June
 
|
  
|
 
* kaz(-kir) trimmed coverage ≥90%
  
* kaz(-kir) trimmed coverage ≥90%
Line 97: Line 134:
 
* tur(-uzb) trimmed coverage ≥80%
  
* tur(-uzb) trimmed coverage ≥80%
 
* uzb(-tur) trimmed coverage ≥80%
  
* uzb(-tur) trimmed coverage ≥80%
  +
|{{Workeval5|3}}
−  
  +
|
  +
|
  +
|-
  ⚫
! 7 !! 29 June - 5 July
  +
|
  +
* get texts for kaz-kir translating
 
|-
  
|-
 
! 8 !! 6 - 12 July
  
! 8 !! 6 - 12 July
 
|
  
|
  +
* clean up kir.lexc
−
* testvoc adjs for all pairs
  
 
|-
  
|-
 
! 9 !! 13 - 19 July
  
! 9 !! 13 - 19 July
 
|
  
|
−
* testvoc numerals for all pairs
 +
* corpus textvoc for kaz-kir
 
|-
  
|-
 
! 10 !! 20 - 26 July
  
! 10 !! 20 - 26 July
 
|
  
|
  +
*
−
* testvoc v.iv for all pairs
  
 
|-
  
|-
 
! 11 !! 27 July - 2 August
  
! 11 !! 27 July - 2 August
 
|
  
|
  +
*
−
* testvoc v.tv categories for all pairs
  
 
|-
  
|-
 
! 12 !! 3 - 10 August
  
! 12 !! 3 - 10 August
 
|
  
|
  +
*
−
* testvoc adverbs for all pairs
  
−
* testvoc misc categories for all pairs
  
 
|-
  
|-
 
!colspan="2" style="text-align: right"|pencils-down week<br />final evaluation<br />11 August - 18 August
  
!colspan="2" style="text-align: right"|pencils-down week<br />final evaluation<br />11 August - 18 August
Line 130: Line 172:
 
* make scripts for:
  
* make scripts for:
 
** getting raw numbers for [[User:Firespeaker/GSoC2014/Workplan|Progress]]
  
** getting raw numbers for [[User:Firespeaker/GSoC2014/Workplan|Progress]]
−
** doing regression tests
 +
** doing regression tests (/learn how to use existing scripts)
 
** [[User:Firespeaker/Cleaning up a tail|guesser]]
  
** [[User:Firespeaker/Cleaning up a tail|guesser]]
 
* get updated corpora for:
  
* get updated corpora for:

Latest revision as of 18:50, 2 July 2014

Primary goals[edit]

  • A production-ready release of kaz-kir
    • Translates kaz→kir and kir→kaz with consistently <10% WER
    • Trimmed coverage for kaz and kir ≥90%
  • A production-ready release of tur-kir
    • Translates tur→kir and kir→tur with consistently <20% WER
    • Trimmed coverage for tur and kir ≥85%
  • A stable release of uzb-tur
    • Translates tur→uzb and uzb→tur with consistently <25% WER
    • Trimmed coverage for tur and uzb ≥80%
  • While bidix size is not built into the goals, the trimmed coverage numbers can be seen as a more relevant proxy for the same basic idea.

Plan[edit]

Schedule[edit]

See GSoC 2014 Timeline for complete timeline.

week dates goals eval accomplishments notes
post-application period
22 March - 20 April
  • apertium-kir to 90% coverage (with kaz-like transducer)
  • apertium-tur to 90% coverage (with kaz-like transducer)
  • apertium-uzb to 90% coverage (with kaz-like transducer)
  • build arsenal of texts with post-edited translations:
    • four 200-word texts in each kaz, kir, tur, uzb
    • four 500-word texts in each kaz, kir, tur, uzb
  • Reworked apertium-tur verb morphology on paper
  • Much better disam in apertium-tur
  • Have a bunch of texts, not many post-edited
  • Need to rework transfer rules for apertium-tur
community bonding period
21 April - 19 May
  • tur-uzb bidix to 7000 stems
  • make some baseline CG for kir, uzb
  • one 200-word kaz-kir text to <10% WER
  • one 200-word kir-kaz text to <10% WER
  • one 200-word tur-kir text to <10% WER
  • one 200-word kir-tur text to <10% WER
  • one 200-word tur-uzb text to <10% WER
  • one 200-word uzb-tur text to <10% WER
  • Implemented reworking of apertium-tur verb morphology
  • Around, but lost time due to end of semester and conferences
1 19 - 24 May
  • one 200-word kir-tur text to <10% WER
  • one 200-word kir-kaz text to <10% WER
  • work on kir CG and lrx
  • Fixed some apertium-tur phonology to go with morphology rework
  • Mostly worked on LREC poster
2 25 - 31 May
  • one 200-word kaz-kir text to <10% WER
  • one 200-word tur-kir text to <10% WER
  • work on kaz CG and lrx
  • work on tur CG and lrx
  • At LREC, worked a lot on apertium-uig and apertium-kaz-uig
3 1 - 7 June
  • one 200-word tur-uzb text to <10% WER
  • one 200-word uzb-tur text to <10% WER
  • work on uzb CG and lrx
  • work on tur CG and lrx
  • Fixed a small handful of transfer issues in apertium-tur-kir
  • Got apertium-uig in a state for others to work with
  • Completed eval for last few weeks
4 8 - 14 June
  • one 500-word kir-tur text to <10% WER
  • one 500-word kir-kaz text to <10% WER
  • work on kir CG and lrx
  • start testvoc nouns for all pairs
  • Brought coverage of apertium-tur-kir on SETimes corpus up by 2%
  • Brought testvoc of apertium-tur-kir on SETimes corpus down from 10.39% to 0.22%
5 15 - 21 June
  • one 500-word kaz-kir text to <10% WER
  • one 500-word tur-kir text to <10% WER
  • work on kaz CG and lrx
  • work on tur CG and lrx
  • continue testvoc nouns for all pairs
  • Brought testvoc of apertium-tur-kir on SETimes corpus down from 0.22% to 0.04%
  • Made and presented poster for Morphology Fest
6 22 - 28 June
  • one 500-word tur-uzb text to <10% WER
  • one 500-word uzb-tur text to <10% WER
  • work on tur CG and lrx
  • work on uzb CG and lrx
  • continue testvoc nouns for all pairs
  • Moving and getting situated week
  • (break for personal reasons)
midterm eval
29 June
  • kaz(-kir) trimmed coverage ≥90%
  • kir(-kaz) trimmed coverage ≥90%
  • tur(-kir) trimmed coverage ≥90%
  • kir(-tur) trimmed coverage ≥90%
  • tur(-uzb) trimmed coverage ≥80%
  • uzb(-tur) trimmed coverage ≥80%
7 29 June - 5 July
  • get texts for kaz-kir translating
8 6 - 12 July
  • clean up kir.lexc
9 13 - 19 July
  • corpus textvoc for kaz-kir
10 20 - 26 July
11 27 July - 2 August
12 3 - 10 August
pencils-down week
final evaluation
11 August - 18 August
  • move pairs to trunk
  • document stuff better on the wiki
  • make the pairs live at turkic.apertium.org

Getting started[edit]

  • make scripts for:
    • getting raw numbers for Progress
    • doing regression tests (/learn how to use existing scripts)
    • guesser
  • get updated corpora for:
    • Uzbek
    • Turkish

Recurring[edit]

  • The end of every week:
  • Constantly:
    • Add good sentences to regression tests
    • Clean up lexc files
      • remove duplicate entries
      • alphabetise sections?
      • add glosses, etc.
Retrieved from "https://wiki.apertium.org/w/index.php?title=User:Firespeaker/GSoC2014/Workplan&oldid=49219"