Blog post dated Sep 2, 2019

Enabling and Using the New Quality Management in Across

More and more users are asking me how to activate and use the quality management of the ATE v7.0.

As I mentioned in my blog post about the six reasons why you should update to ATE v7.0, it would make sense to familiarize yourself with the quality management of the ATE v7.0 at an early stage. Meanwhile, all quality criteria have been set up in the ATE v7.0 Gilded Sapphire, so this is a good time for the transition. The transition is not really a trivial matter, and some preliminary work needs to be done, but once this has been done, you will be able to use the Quality Management v7.0 whenever you wish.

An Article by

Andreas Rodemann
Trainer Across Translator Edition

In this article, you will learn with which settings you can optimally use the new Across quality assurance.

Necessary Settings

The following settings apply to the standalone mode only. For the Offline Client mode, the necessary settings will be configured by your customer. First, go to the System Settings and configure your QM template (QST) there.

The following settings apply to the standalone mode only. For the Offline Client mode, the necessary settings will be configured by your customer.

You can either use the "Default" template or create a new one (which you can name as you wish) by clicking the "Actions" button. I will call it "MyQM_v7" here. Click the small arrow in front of each criterion to access the settings you can configure for the respective criterion. For most criteria, the options are limited to the "Enabled" and "Mandatory" settings, which you can usually leave at the default settings. However, some additional options are available for the following criteria. I will only explain the meaning of criteria and options that are not self-explanatory.

Criterion Additional options Enabling recommended
Spacing (checks for spaces between words) Check consecutive white spaces Yes
  Ignore if same sequence is used in source Yes
  Check tabs Yes
  Check bounding spaces (checks for white spaces at the beginning and end of segments) Yes
Data/time/numbers None Yes
Formatting Check formatting order No
  Warn if rich text formatting is used in target Yes
Identical segments (checks if segments in source and target text are identical) Ignore paragraphs with characters fewer than XX (10; if you wish, you can test what number of characters is appropriate)
Same source text – different translation text None Yes
Brackets None Yes
Leerer Empty target None Yes
Placeables* Check placeables order Yes
  Check placeables formatting Yes
Private use area characters** None No, if client does not use such characters
Spelling Ignore if same word is used in source Yes
  Ignore all-caps words Yes
  Ignore words with numbers Yes
End punctuation None Yes
Terminology Use stemming*** No
  Only consider released terms Yes, if client-specific terminology is important
  Only consider terms of the same relation as the current project Yes, if client-specific terminology is important
Display text bounds None No, if no display text is translated
Translation memory Check for ##NO_MATCH## placeholders Yes
  Check for unchanged fuzzy matches Yes
  Check for unused 100% matches Yes
Capitalization Warn about missing initial capitalization without taking the source into consideration Yes
Segment count None Yes
Different source test – same translation text None Yes
XML validity None Yes, if XML texts are translated

*Placeables are gray fields containing information.

**Private Use Area characters are characters in the Unicode character map that can be defined by the user without any restrictions.

***Stemming is the function that separates terminology into a stem and an ending in order to suggest terminology. This feature is useful for translating, but it shows too many errors when checking the terminology.

Of course, you can also create additional QM templates, e.g. if you need to check certain criteria for a particular customer, but not for another.

Once you have set up your QM template, go to "Project Setting Templates" in the system settings.

There, create a new project settings template (PST) and name it e.g. "QM_v7". Enable the function "Quality Management v7.0" for this PST and select the template you have just set up under "QM template". If necessary, adjust the other PST settings, e.g. "Select system attributes", "Storing Settings, "Penalty Settings", and "Pre-translation Settings" on the "crossTank" tab, and "Maximum word combination length" on the "crossTerm" tab.

Of course, you can also create additional QM templates, e.g. if you need to check certain criteria for a particular customer, but not for another.

When you create a new project, select the PST "QM_v7" on the second page of the Project Wizard at the top left under "Project settings".

You are now ready to use the quality management of the ATE v7.0. You can of course also activate QM v7.0 for the default PST, but in this case, you would have to create a new PST in order to be able to use the Quality Management v6.3.

Therefore, I would only recommend activating QM v7.0 in the default QM template if you want to use QM v7.0 permanently instead of selecting the PST each time in the Project Wizard.

Quality Management v7.0 in crossDesk

After you have completed your translation, you can run quality checks. To do so, go to the QM tab in crossView and click "Check All".

Across will collect and list all potential errors. You can then process this list one item after the other. Use the arrow buttons or the keyboard shortcut Ctrl+Shift+Down to jump to the next or previous error (Ctrl+Shift+Up). The eye icon allows you to hide or unhide errors that you have marked as "Ignore". The funnel icon allows you to filter criteria to process only one criterion at a time. This is similar to the QM batch mode, which is no longer available in QM v7.0. When you select a paragraph in the contextView, the QM list will automatically highlight all errors for the respective paragraph.

Similarly, when you select an error in the list, the respective paragraph will automatically be activated. Depending on the criterion, you can handle the displayed error in different ways. For some errors, you will have to correct the error yourself or select "Ignore" (Ctrl+Shift+I). You also have to "Ignore" errors that you do not correct. This is not so important in the standalone mode, since you are the one in charge. However, you will have to do this for the mandatory criteria in the Offline Client, as otherwise you will not be able to finish the task.

In my opinion, it would be best to expressly "Ignore" all errors that you do not correct in the Offline Client mode, i.e. not only the mandatory ones. In this way, your customer will notice that you have seen the error and dealt with it. For other criteria, you will also see the option "Apply" (Ctrl+Shift+Enter). I find this feature very convenient, as Across will automatically correct the particular error without you having to find and correct the error manually in the Target Editor.

The "Spelling" criterion gives you additional options: You can add the word to the dictionary by clicking "Add to Dictionary", filter it with "Similar", or apply a suggested word, in which case the error will be corrected automatically without you having to intervene.

When you select a paragraph in the contextView, the QM list will automatically highlight all errors for the respective paragraph.

Conclusion

Now that I have worked on several projects using the new quality management functionality, I must say that I like it. In particular, I like the automatic error correction, the possibility to control the QM almost entirely with the keyboard, and the consolidated range of criteria. Of course, it will take some time for you to get used to the new QM, and so far, I haven't tested whether the new quality management is faster than the old one. I will need some more projects to find out. But my intuition is that it is faster simply because I no longer have to find every single mistake in the Target Editor and fix it manually.

About the Author

Andreas Rodemann is a graduated translator with Chinese and English as working languages. He has been working as freelance translator with Across since 2009. Besides this, Andreas Rodemann has been moderator for many years and since 2014 owner of the “Übersetzer- und Dolmetscher-Lounge” (Translators and Interpreters Lounge) at XING where he is assisting fellow translators and interpreters with words and deeds far beyond Across topics.

Website: AR-Übersetzungen