How does the translation memory work?

What is a translation memory?
In order to improve both the speed and consistency of our translations, we use translation memories. A translation memory is a database that stores "segments", which can be sentences, paragraphs or sentence-like units (headings, titles or elements in a list) that have previously been translated, in order to aid human translators.

Text analysis
Every text is automatically analysed for repeat segments. You can find this analysis in the assignment overview, where you can also see the earnings.


The translation process
1: Start translating

2: If (part of) the text matches prior translations for this client, this translation will appear as a suggestion (figure 1). You can use this translation simply by clicking on it or by using the arrow keys and the enter key on your keyboard.

3: Only use the suggestion if you feel it's appropriate or if instructed to do so by the client.

4: After you've chosen the suggestion you can edit it freely.


What happens behind the scenes

  • A translation memory is created for every client and for every unique language pair.
  • During the translation process, after sending a page (figure 2), the translation memory will be amended.
  • If the next translation page contains any segments that have already been translated before for this client, you will be shown suggestions based on the translation memory.
  • This results in a translation memory for each client based on all the corresponding assignments.
  • If multiple assignments in the same language pair have been completed for the client, you may also receive suggestions from those assignments. This included assignments that have been translated by other translators.


Figure 1



Figure 2


Have more questions? Submit a request


Powered by Zendesk