Best way to perform QA of translation memory file (tmx)
Thread poster: Dimmo Petrov

Dimmo Petrov
Local time: 21:16
English to Bulgarian
Aug 22, 2013


Can you please share what is the best way to perform QA of translation memory file (tmx)?

Thank you in advance!



Heartsome Support
Local time: 03:16
For your reference Aug 23, 2013

First you have to find a good tool for running QA of translation memory file. There are several free and commercial tools.
 Heartsome TMX Editor
 Olifant
 vTMXEditor
 Xbench
 QADistiller

Here some suggestions for your reference:
1. Find and delete duplicate sentences, leave one sentence only.
2. Find the segments that have same source with different translations, keep one translation and delete the rest.
3. Check and correct the terms in the TMX
4. Delete rogue codes
5. Edit the attribute/property information, such as Modified by, Client, Job# etc.
6. Convert the TMX to TMX level one, i.e. delete all formatting information in the TMX. This usually could make your TMX cleaner, and could raise the match rate. But this may not apply for all situations. You can make your own choice.

The latest version of Heartsome TMX Editor could do all above either manually or automatcially. Once it is released, you can try it.


wilhelm_zwo (X)
Local time: 20:16
German to Dutch
CafeTran as a great TMX editor Aug 23, 2013

Besides a magnificent CAT tool, CafeTran (80 Euro) also is a great TMX editor. In TMX edit mode you can use the following features:

QA menu:

And, from the Task menu:

Remove target segments the same as source
Clear those target segments in the active project that are identical to the corresponding source segments. Use this to clear pre-segmented Trados TTX files.

Delete TMX units with duplicate source segments
Remove all segments from a TMX project where source segments have been stored with multiple translations (either identical or different). The newest TU (the one with the highest line number in the TMX file) is kept.

Delete filtered TMX units
Filter on a certain string or regular expression or QA condition in source and/or target segments. Then remove all matching segments from the TM. Very powerful. Make a backup of your TMX file first. Errare humanum est.


SDL Community  Identity Verified
United Kingdom
Local time: 20:16
Member (1970)
Another way... Aug 23, 2013

... would be to simply open the TMX into a CAT and use the QA tools in the CAT to tidy it up. If you are able to add the it as a TM at the same time you will get all the matches and should be able to correct them as you work.

If you have Studio you can convert the TMX to an SDLTM, then create a Project using XLIFF files containing the contents of the SDLTM, and work that way. Finally exporting the QA'd TM to TMX at the end. There is a blog article describing this very process here:

I do have a filetype for TMX that will go on the OpenExchange very soon, but using the method above you don't really need it, and as a TM can be pretty large it may make sense to break it down into bite sized chunks anyway.




Dimmo Petrov
Local time: 21:16
English to Bulgarian
Thank you! Aug 26, 2013

Your help is much appreciated!



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

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

Best way to perform QA of translation memory file (tmx)

Advanced search

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 »
Déjà Vu X3
Try it, Love it

Find out why Déjà Vu is today the most flexible, customizable and user-friendly tool on the market. See the brand new features in action: *Completely redesigned user interface *Live Preview *Inline spell checking *Inline

More info »

  • All of
  • Term search
  • Jobs
  • Forums
  • Multiple search