Our Post-Editing Process
To ensure a consistent level of quality, human intervention should occur at several points of a post editing process:
1) Pre-editing refers to work done before importing the text into the translation system. This includes
- Identifying what should be translated, and excluding what should not;
- Cleaning up and simplifying text formatting (such as data tables, information boxes and graphics);
- Correcting basic spelling, grammar and punctuation errors (machines can be confused by number and word combinations, use of punctuation characters between words and incorrect spelling).
2) Post-editing refers to the work of a translator reviewing the machine translation.
3) Post-editing Review refers to a client or internal review of the post-editing, including comments and questions posed by the translator.
What is Adaptive Machine Translation?
Adaptive Machine Translation, or Adaptive MT, is a technology that allows a translation system (such as Lilt) to learn as soon as it receives corrective input from a human user.
As the user corrects the automated translation (a process called post editing), the system learns from the feedback and immediately integrates it in its rules to avoid similar mistakes.
Is Quality Machine Translation Possible?
Like any project, the quality of your machine translation project will depend on many factors, such as
- Complexity of the subject matter
- Completeness of your glossary and/or translation memory (TM)
- Quality of your pre-editing
- Qualifications of the post-editing translation, review and quality assurance team