HOME2

    Build English-Lao-English Online Dictionary to Google Translate

Our Group LaoEnGT (http://groups.google.com/group/laol10n-laoengt) would like to invite you if you are a person with Lao and English language background to come and help create an English-Lao-English dictionary and Translation Memories (TM) online to Google Translate so that Lao language would be promoted in Google Translate as 'available' or 'supported'. This would facilitate translation from English to Lao and vice versa, and help with communication and information exchange. Currently only major languages are available, but not minor ones like Lao, or Khmer for examples. For Lao language it's up to Lao speaking people to make it happen. 

There is an tool provided by Google, called Google Translator Toolkit (or GTT) for use to build such facility, and its web link is ...

Hint: Select 'Translator Toolkit' at the left of the page, after you sign in.

Note LaoEnGT group is part of a larger movement for development of Lao language for use in the Internet and computers. See ...

 http://groups.google.com/group/laol10n/about?hl=en 

 
Current status of play regarding translation

Currently (or as of September 2010) if one tries to translate Josh's words (Senior Software Engineer at Google Translate) using GTT for translation from English to Lao one gets 2% translated, just the word  'Hello,' to  'ສະບາຍດີ'. The rest remains untranslated. See screenshot below.

    

 
If one does the same but select 'Thai' as translation target lauguage one gets a substantially better result. See below. This is because Thai language in GTT has enough support data in the system, such as 'Glossary', 'Computer Translation' and 'Global Shared Translation Memories (or TM)' features. See online help in GTT for more information.  
 
 
 
Proposed Solution for Lao language
 
 In order to achieve our stated goal we need to do the following:
-  build English-Lao 'Glossary', and use it as initial support for translation activities,
-  build 'English-Lao Shared, Global Translation Memories (or TM)', by virtue of using GTT to do translation from English to Lao with the support of the Glossary mentioned above. Translation Memories are saved translated sentences or lines of text resulted from ongoing translation exercises from anyone participating. This (default) TM file will grow gradually in GTT repository. GTT operates by recommending matched TM units for use in translation when found.
-  build English-Lao 'Computer Translation'. This feature will be invoked when there are no matched TM unit found. Matched TM units are preferred because they are more correct in translation context.
-  optionally save translated documents in a common location for easy access to Lao communities.
 
For detailed explanation of features mentioned above see GTT's Online Help
 
GTT can be used to help with general translation exercises even if there is only just a Glossary for support. However it would be much easier if the other 2 features, 'Computer Translation' and 'Translation Memories' are also present. Note 'Translation Memories' takes precedence over 'Computer Translation'.  
Given above the priority would be 'build the Glossary file'. Note the other two supporting features mentioned can be built later or at the same time.
 
 
Build English-Lao 'Glossary'

We currently have multiple Lao-to-English word pair files that was prepared and nearly ready for creating GTT's English-Lao Glossary file. They are available and stored at

http://sites.google.com/site/eng2lao/home/quality-control-phase

  
They just need some minor review and correction. This work process is described in section
'Word Pair Quality Control (WPQC) and Peer-review'. See http://sites.google.com/site/eng2lao/home/quality-control-phase.
The output from above will be used to create English-Lao Glossary Online file.
 

For current release of the online version see  ...

http://translate.google.com/toolkit/gl?gid=g_7bac6b14-168e-4c53-bcee-9921870dad86&hl=en#

For corresponding non-online version see file 'English-Lao-English Glossary Ver 0.1 (2011May).txt'   in folder  'Lao-English Glossaries', located at the bottom of web page ...

http://sites.google.com/site/eng2lao/home/quality-control-phase 

We need volunteers for the task, please email us (to luan_vannithone@hotmail.com) or
Click below and fill in this Volunteer Application form online ...
 
The Level of work progress can be ascertained by the percentage completed. See below for example.

List of volunteers ...

Person                                            Task (Word pair file)                     % completed    Date startedas

----------------------------------------------------------------------------------------------------------------------------------------------------------------- 
Bounluang Vannithone                      LaoEng04.txt                                2%                  25-May-2010
Chittakone Sengchansavang             LaoEng03.txt                                 3%                   4-Jun-2010
Outhong Phongsavath                       LaoEng02.txt                                1%                   4-Jun-2010
Sourith Kayasith                               LaoEng06.txt                                1%                   4-Jun-2010
Souriya Mekdarasouk                       LaoEng05.txt                                1%                   4-Jun-2010
Thongrith Phoumirath (Dr)                LaoEng01.txt                                 1%                   4-Jun-2010

..

(Approximately 28+ files are available, each contains approx 500 Lao-English word pairs. They are based on the National University of Laos (NUOL) Spot Dictionary version 0.1 (June 2010)

For complete and up to date information on current progress see Progress Report page, section 'Current Progress Status" .http://sites.google.com/site/eng2lao/home/progress-report
 
Below are sample resources you will find useful.
- A Lao 'corpus' type word search http://sealang.net/lao/dictionary.htm  by CRCL and the University of Wisconsin-Madison
- A free Windows based Lao-English-Lao dictionary program from Soutsaka College http://www.simt.edu.la/Students%20work.html 
 
For linguistic type information see section 'Questions and Answers' ... https://sites.google.com/site/eng2lao/home/-questions-and-answers
 
 
How is Glossary used for translation in GTT ?
 
See below, for example, the system tries to translate text segment or sentence "Just about every day we hear from a user, or potential user, asking us". It can not find a match from the TM, therefore the text substitution doesn't take place however the text is placed in the 'translation box' at the right hand side for you to work on. In order to make life easier it also marks or highlighted those words found in the Glossary, for example, ABOUT, DAY, A, and ASKING, to indicate equivalent Lao words you can select (copy and paste) for translation. The list of Lao words or synonyms appears at the bottom right corner.    
 
 
 
Once the text seqment in focus is translated and saved (after clicking 'Next' in the Translation box') the English-Lao TM unit will be created and saved in the Global, Shared Translation Memory repository or custom Translation Memory (if specilied), thus becomes readily available to anyone who later want to translate the same or 'very similar' text segment to Lao.
  
Note the Glossary file can be shared to group members and selected for use in translation from the 'Use Glossary:' dropdown field, as shown in sample screenshot below.
 
 
 
 
 
How to measure project work progress regarding building English-Lao Translation Memories (TM)?
 
The bigger the pool of English-Lao Translation Memories (TM) becomes the easier GTT is for doing translation to Lao, and the closer we will approach the end goal - Lao language as an 'available' or 'supported' language in Google Translate.
However there is an important question that needs to be addressed - How does one go about measuring project work progress regarding building English-Lao Translation Memories (TM)?.
 
Since GTT doesn't quite provide a mean to show or display detail/statistics on their (default) Global Shared Translation Memory file, we need to devise a mean for it.
To that end one proposed mean is to use and share an 'custom collaborated TM file'. GTT does provide support for its creation and dowloading of collaborated and shared TM file. The number of TM units can be counted and reported on after downloading.
An added benefit of this custom TM file is its ability for use in reversed translation, i.e. from Lao to English, after a minor change, swapping the order of the TM language pair. Sharing and collaborating on custom TM file is similar to using Glossary. See section
'How is Glossary used for translation in GTT ?'
 
 
In Summary...

Once this endeavor is completed (or almost) the task of translating between English and Lao should be much easier for anyone, thus enhancing communication and facilitating availability of technical and other knowledge in Lao language. 

For more info see discussion thread http://groups.google.com.au/group/laol10n/browse_thread/thread/6efb39a569fe043f?hl=en

>

[ We are considering accepting financial contributions from those who want to help, but do not have spare time to do so. The fund will be used to pay unemployed Lao persons (or any suitable person with English and Lao language knowledge skills) and who are looking for something worthwhile to do. To contribute click here http://sites.google.com/site/eng2lao/home/z-contribution-1. ]

>

As of 6-June-2010 I'm happy to announce we have 7 inaugural members/ volunteers to help drive the project progress. See list of names in Progress Report page, section 'Volunteer or member List'.

http://sites.google.com/site/eng2lao/home/progress-report

A big 'Thank-you' to these individuals.
 

 >>>>>

Luan (or Bounluang) Vannithone  3-May-2010

Updated on:          15-Aug-2011