Difference between revisions of "User:Rafi kamal/Coding Challange GSoC 2014"

From Apertium
Jump to navigation Jump to search
(Created page with "As part of the [http://wiki.apertium.org/wiki/Ideas_for_Google_Summer_of_Code/Make_a_language_pair_state-of-the-art coding challenge], I've done the following: * Installed Ap...")
 
Line 1: Line 1:
As part of the [http://wiki.apertium.org/wiki/Ideas_for_Google_Summer_of_Code/Make_a_language_pair_state-of-the-art coding challenge], I've done the following:
+
As part of the [http://wiki.apertium.org/wiki/Ideas_for_Google_Summer_of_Code/Make_a_language_pair_state-of-the-art coding challenge] suggested by Francis, I've done the following:
   
 
* Installed Apertium and bn-en language pair from SVN
 
* Installed Apertium and bn-en language pair from SVN
 
* Translated four 500 page articles from Bangla to English using Apertium
 
* Translated four 500 page articles from Bangla to English using Apertium
 
* Postedited these four articles to create a reference translation
 
* Postedited these four articles to create a reference translation
* Used apertium-eval-translator to calculate word error rate (WER) and position independent word error rate (PWER) of existing translation.
+
* Used <code>apertium-eval-translator</code> to calculate word error rate (WER) and position independent word error rate (PWER) of existing translation
  +
  +
Here is the evaluation result of existing bn-en translation:
  +
  +
{|class="wikitable"
  +
|-
  +
! Article Number
  +
! WER
  +
! PWER
  +
|-
  +
| 1
  +
| 94.35 %
  +
| 89.01 %
  +
|-
  +
| 2
  +
| 93.39 %
  +
| 85.89 %
  +
|-
  +
| 3
  +
| 94.54 %
  +
| 87.68 %
  +
|-
  +
| 4
  +
| 94.80 %
  +
| 89.95 %
  +
|}
  +
  +
The translated articles can be found [https://github.com/rafi-kamal/Apertium-Coding-Challenge here].

Revision as of 14:54, 28 March 2014

As part of the coding challenge suggested by Francis, I've done the following:

  • Installed Apertium and bn-en language pair from SVN
  • Translated four 500 page articles from Bangla to English using Apertium
  • Postedited these four articles to create a reference translation
  • Used apertium-eval-translator to calculate word error rate (WER) and position independent word error rate (PWER) of existing translation

Here is the evaluation result of existing bn-en translation:

Article Number WER PWER
1 94.35 % 89.01 %
2 93.39 % 85.89 %
3 94.54 % 87.68 %
4 94.80 % 89.95 %

The translated articles can be found here.