Post-editing for machine translations

Post-editing refers to the process whereby machine-translated texts are checked by a proofreader before being published.
This service was developed for companies that use machine translation solutions for their (primarily internal) international communication.

Even with careful linguistic and technical preparation, machine translations are not usually completely free of errors. Proofreading by a proofreader trained and experienced in the field of post-editing is therefore highly recommended.

One advantage of this process is that it can also include an evaluation of the quality of a machine translation. This can allow the company to evaluate how suitable the data in the machine translation engine, known as the corpus, has been in the past and how it is developing.

The quality and reliability of the corpus is also continuously improved through the correction of incorrect or inappropriate segments.

The precise structure of a post-editing process can be arranged on an individual basis.

tsd would be happy to advise you on the topic of post-editing for machine translations or to plan individual solutions for your company. Simply contact us by telephone on +49 (0) 221 9259860 or write us an email at

  1. tsd’s project manager receives the machine-translated or partially machine-translated files.
  2. They forward them to the proofreader, who reviews and corrects the translation. Depending on the workflow, the proofreader works directly in the machine translation tool, in the translation management system or in the text files. Whether or not corrections are entered into the corpus is decided on an individual basis.
  3. When they have finished their work, the proofreader sends the project manager the corrected files and, if relevant, an evaluation of the quality of the machine translation (BLEU score).
  4. The project manager delivers the documents to the customer.


How was the machine translation created? What data was used to create the translation engine? What is the corpus based on?

Are there any special work instructions?

Should our proofreaders work directly in the software or in the document?

Who will enter the corrections into the corpus after processing?

Would you like an evaluation of the translation quality? If so, what standards should this evaluation be based on? Should any developments in the translation quality be documented?