Pretranslating documents

Pretranslating a document means pre-populating its target segments according to certain rules. It is useful:

  • When your document has a lot of translation memory matches. This will prevent you from having to click through and confirm each segment found in the TM.
  • When your document has number-only segments. Smartcat will insert numbers and even convert their formatting according to language rules.
  • For machine translation post-editing. Smartcat will fill every segment with the machine translation of the source text, for a translator to edit and confirm.

In Smartcat, you can set up pretranslation either when creating a project or at any time after that by going to the Pretranslation tab of the project page. Note that the same pretranslation rules are applied to all existing and future project documents — so you don’t need to configure them for individual documents.

Pretranslation rules are applied in steps, and for each rule you define:

  • What to insert: translation memory matches, numbers, or machine translation.
  • For translation memory matches, TM name, match percentage and minimum word count needed for a match to be inserted (this prevents you from pre-translating very short segments that could be translated differently depending on the context - for example, image captions).
  • Whether to confirm the segment after insertion. 

Note: Confirming segments will result in the segments being locked for freelancers assigned to this step and the previous steps. For example, if you confirm a segment at the editing stage, the freelancer assigned to the translation stage will not be able to make any changes - the segments will be locked. You would also not need to pay the freelancer for these segments. While the editor could make changes, since the segments are already confirmed, that freelancer would not get paid for the work. 

By defining different rules and applying them in different orders, you can achieve flexible pretranslation scenarios.

For example, you could specify the following steps:

  1. Insert and confirm numbers and confirm them at the Editing stage.
  2. Insert and confirm 101% matches (context matches) longer than 2 words from your primary TM and confirm them at the Editing stage. 
  3. Insert but don’t confirm 75% matches from your primary TM.
  4. Insert but don’t confirm 100% matches from another (say, domain-specific) TM.
  5. Insert but don’t confirm machine translation.

pretranslation-rules.png

As a result, translation costs and the time needed to complete the projects are both greatly reduced. 

See also: Comparison of pretranslation features in MemoQ and Smartcat.