SDL Trados 2007 Workbench Analyse Tool
Thread poster: Gabriel Mendoza

Gabriel Mendoza  Identity Verified
Japan
Local time: 04:50
Japanese to English
Oct 19, 2010

Hi. This is a follow-up on my post last July. You can find the old post here:

http://www.proz.com/forum/sdl_trados_support/175858-translators_workbench_analyse_tool.html

It's still pretty much unresolved. Just to give a recap:

I'm using SDL Trados 2007 and I want to analyze a translated file to get statistical information on repetitions, fuzzy matches, and such. However, when I run the Analyse command in Workbench and analyze the target (translated) document, analysis results only show counts in number of characters and not in number of words. I reckon this is because the TM I'm using has Japanese as the Source language. Is there any way to configure the tool to count words instead of characters?

My language pair is Japanese to English. I need to analyze the target (translated) document for invoicing purposes. The outsourcer pays me for every target English word I produce and not per character. They pay me at full rate per target word except repetitions and 100% matches, for which they pay less. That is the reason why I HAVE to analyze the target document and have the results in number of words.

Any help will be much appreciated. Thank you in advance.


 

Jerzy Czopik  Identity Verified
Germany
Local time: 21:50
Member (2003)
Polish to German
+ ...
Create a TM English to Japanese Oct 19, 2010

Use that one to analyze your target.
After you've done the analyze update the TM with an export from your JP-EN TM.
That way with each job you will analyze the file against the TM from previous job.
Analyzing English text against a Japanese TM does not make much sense.


 

Gabriel Mendoza  Identity Verified
Japan
Local time: 04:50
Japanese to English
TOPIC STARTER
Doesn't seem to work or I might be doing something wrong Oct 20, 2010

Hi Jerzy,

Thank you for the reply. Here is what I did:

(1) Opened my existing JPN-ENG TM and exported it to a TMX file.
(2) Created a new ENG-JPN TM.
(3) Opened the Analyse tool in the new TM and analyzed the target document. (It returned some repetitions which puzzles me because the TM is supposed to be new and empty, right? :S).
(4) Closed the Analyse window and imported the TMX file in number (1) into the new TM.
(5) Opened the Analyse tool again in the new TM and analyzed the target document. It returned a large amount of repetitions which is totally inaccurate considering that the document only has 2 pages and virtually has no repetitions.

So it seems I'm back to square one. Or, I could be doing something wrong.icon_frown.gif


 

Jerzy Czopik  Identity Verified
Germany
Local time: 21:50
Member (2003)
Polish to German
+ ...
Repetitions do appear also for an empty TM Oct 20, 2010

So no wonder - repetition is not connected to any TM, as this is simply a sentence appearing in a document more than once.

Please do the following:
- Export your JPN-ENG TM to TXT (not TMX).
- Create a new ENG-JPN TM (please pay attention, that the English flavour is the same, eg. EN-US or EN-UK)
- Analyse the TARGET (the file in English) file against the empty ENG-JPN TM
- Import the TXT and analyse the file again. This should now deliver only high matches and repetitions.

Otherwise I do not really understand what's wrong. However my experience with languages like JPN is zero, so it might well be, that WB behaves there differently from what I know for languages with latin characters.


 

Gabriel Mendoza  Identity Verified
Japan
Local time: 04:50
Japanese to English
TOPIC STARTER
Same results Oct 20, 2010

Followed your instructions. Got the same results. Got figures across all ranges and still "100% match" dominates.icon_frown.gif

I'm at wit's end. Trados didn't even count the source characters correctly when I tried analyzing the source document. Sigh...

Do you happen to know of any other analysis tool out there that I can use?

Thanks in advance.


 

Jerzy Czopik  Identity Verified
Germany
Local time: 21:50
Member (2003)
Polish to German
+ ...
Sorry, I do not know Oct 20, 2010

If you would send me the translated bilingual file I could try that for you.

My mail is info at tts - td . com (obviously without spaces)


 

Gabriel Mendoza  Identity Verified
Japan
Local time: 04:50
Japanese to English
TOPIC STARTER
Test Results Oct 20, 2010

It's interesting to note that from the results of my tests (using just the original TM), the Analyse tool is actually counting English characters (without spaces) in the target document. And it's accurate. The segment count is also very accurate for each percentage.

What I simply want it to show is not the number of English characters, but the number of English words.


 

Jerzy Czopik  Identity Verified
Germany
Local time: 21:50
Member (2003)
Polish to German
+ ...
Analysis results are always words for latin letter languages Oct 20, 2010

Still can't get your point on that - the analysis does give you the number of segments, words, characters without spaces and percentages.
Do I miss something here?


 


To report site rules violations or get help, contact a site moderator:


You can also contact site staff by submitting a support request »

SDL Trados 2007 Workbench Analyse Tool

Advanced search







WordFinder Unlimited
For clarity and excellence

WordFinder is the leading dictionary service that gives you the words you want anywhere, anytime. Access 260+ dictionaries from the world's leading dictionary publishers in virtually any device. Find the right word anywhere, anytime - online or offline.

More info »
SDL Trados Studio 2017 Freelance
The leading translation software used by over 250,000 translators.

SDL Trados Studio 2017 helps translators increase translation productivity whilst ensuring quality. Combining translation memory, terminology management and machine translation in one simple and easy-to-use environment.

More info »



Forums
  • All of ProZ.com
  • Term search
  • Jobs
  • Forums
  • Multiple search