Translation showing 0% matches and verification reporting segment not edited - how?
Thread poster: Francesca Collodo
Francesca Collodo
Francesca Collodo  Identity Verified
United Kingdom
Local time: 12:18
German to English
+ ...
Jan 13, 2020

Hi all, I'm a bit baffled here and hoping you can help.

I do a lot of reviewing of test translations for a company, today I've come across something I've never seen before.

There are multiple segments reported as "Target confirmed without prior editing". These segments show a TM match result of "0%". So the target was clearly auto-populated - but from where? The content is not in the TM.

There is a TM included with the project. Rather than reporting that th
... See more
Hi all, I'm a bit baffled here and hoping you can help.

I do a lot of reviewing of test translations for a company, today I've come across something I've never seen before.

There are multiple segments reported as "Target confirmed without prior editing". These segments show a TM match result of "0%". So the target was clearly auto-populated - but from where? The content is not in the TM.

There is a TM included with the project. Rather than reporting that the match was found in this named TM (which it wouldn't be, as it is supplied empty, as it's a test translation), when I hover over the match it says "Origin: Translation Memory" and "Provider: Translation Memory".

There are also several other similar segments, which say "Origin: Auto-propagated" and "Provider: Propagated". These segments appear only once in the file.

How has the translator achieved this? Obviously, I am trying to rule out the use of MT here.

Any advice gratefully appreciated!
Collapse


 
Rossana Triaca
Rossana Triaca  Identity Verified
Uruguay
Local time: 09:18
English to Spanish
sdlxliff files are easily editable Jan 13, 2020

Trados' bilingual files are simply a flavor of XML, so it's very easy to open them up in a text editor (Notepad) and change any and all segments into having whatever status and origin you prefer, or to hide any traces of MT used. You can modify absolutely everything, including the "origin", "conf" status, match "percent", "created_on", "created_by" and any other info you can actually see in Trados.

Most of the time however, when you have segments with, let's say, "unconventional" or
... See more
Trados' bilingual files are simply a flavor of XML, so it's very easy to open them up in a text editor (Notepad) and change any and all segments into having whatever status and origin you prefer, or to hide any traces of MT used. You can modify absolutely everything, including the "origin", "conf" status, match "percent", "created_on", "created_by" and any other info you can actually see in Trados.

Most of the time however, when you have segments with, let's say, "unconventional" or conflicting statuses, it usually means that the translator was CAT hopping, i.e., using a different CAT or third-party tool to process and translate the sdlxliff files. This is definitely common if they like some other environment better and/or have a ton of resources already for a different tool and they are proficient enough to guarantee that processing them outside of Trados yields the same results.

So, in short, you can't rule out MT programmatically and you'll have to determine linguistically if something is amiss (although, if you can't tell by the end product, I personally feel this is a non-issue). However, I understand the trust between the vendor and the client is paramount and the translator would be shooting him/herself in the foot by using MT if it's specifically against a client requirement, not to mention it's ethically wrong to agree to some terms you have no intention to keep of course.

[Edited at 2020-01-13 16:30 GMT]
Collapse


Francesca Collodo
 
Francesca Collodo
Francesca Collodo  Identity Verified
United Kingdom
Local time: 12:18
German to English
+ ...
TOPIC STARTER
Mystery solved :) Jan 13, 2020

Rossana, thank you! That makes sense.

It is in the client contract that MT shall not be used (sensitive information they don't want passed to external servers). Though I do agree - I have no objection to a good result achieved with careful use of MT. But I do understand the client's requirement.

I had just never seen this before. Now it makes sense, how this may have happened!


 
AnnaSCHTR
AnnaSCHTR  Identity Verified
United States
Local time: 07:18
English to Czech
+ ...
A different TM? Jan 15, 2020

I was just wondering if it could be caused by the translator's using a different TM? I know some translators have their huge master monster TMs, or TMs for specific subject matters. Or would it be displayed differently?

 
mikhailo
mikhailo
Local time: 15:18
English to Russian
+ ...
re Jan 15, 2020

Francesca Collodo wrote:

Hi all, I'm a bit baffled here and hoping you can help.

I do a lot of reviewing of test translations for a company, today I've come across something I've never seen before.

There are multiple segments reported as "Target confirmed without prior editing". These segments show a TM match result of "0%". So the target was clearly auto-populated - but from where? The content is not in the TM.

There is a TM included with the project. Rather than reporting that the match was found in this named TM (which it wouldn't be, as it is supplied empty, as it's a test translation), when I hover over the match it says "Origin: Translation Memory" and "Provider: Translation Memory".

There are also several other similar segments, which say "Origin: Auto-propagated" and "Provider: Propagated". These segments appear only once in the file.

How has the translator achieved this? Obviously, I am trying to rule out the use of MT here.

Any advice gratefully appreciated!


ALL CATs support SDLXLIFF. Files could be translated in any of them.


 
Roy Oestensen
Roy Oestensen  Identity Verified
Denmark
Local time: 13:18
Member (2010)
English to Norwegian (Bokmal)
+ ...
Sorry, but not true Jan 19, 2020

mikhailo wrote:
ALL CATs support SDLXLIFF. Files could be translated in any of them.

All sensible CATs support SDLXLIFF, including memoQ and Dejavu, support the xliff format. But some CAT tools are much more insulated. A lot of cloud based CAT tools that I have experience with, do not support the xliff format, and the same for some offline CAT tools. Most translators I know, regard these as inferior to the most common tools like Studio and memoQ, though.

Roy


 


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


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

Translation showing 0% matches and verification reporting segment not edited - how?







TM-Town
Manage your TMs and Terms ... and boost your translation business

Are you ready for something fresh in the industry? TM-Town is a unique new site for you -- the freelance translator -- to store, manage and share translation memories (TMs) and glossaries...and potentially meet new clients on the basis of your prior work.

More info »
CafeTran Espresso
You've never met a CAT tool this clever!

Translate faster & easier, using a sophisticated CAT tool built by a translator / developer. Accept jobs from clients who use Trados, MemoQ, Wordfast & major CAT tools. Download and start using CafeTran Espresso -- for free

Buy now! »