The world of translation is becoming increasingly interactive. In addition to translation memory and terminology, the translator now often has access to translation suggestions from the machine. How can man and technology work together successfully? What is post-editing and what skills does a post-editor need to bring or acquire? This is what we want to explain in this blog post.
Post-editing – what is it and why do we need it?
Post-editing is – to put it simply – the improvement of machine translation results.
It is the process by which a machine-generated translation is checked and adjusted by a human being so that it reaches an acceptable quality. A distinction is made between light and full post-editing:
Depending on the quality requirements and the type of text, you may opt for light post-editing, in which only the very necessary corrections are made, or full post-editing, after which it may no longer be possible to distinguish whether the translation was produced by a human or a machine.
What makes a post-editor?
Many of the skills and characteristics that are a prerequisite or common for a translator can also be found in post-editors. But what distinguishes a post-editor from a translator? Ideally, the post-editor has knowledge of machine translation and understands how it works.
Making quick decisions
In addition, a post-editor should also always keep an eye on the profitability of his work. Especially when a post-editing job is not about flawless, but “only” comprehensible texts, the editor has to cut back on quality, put aside his perfectionism and make quick decisions about what needs to be corrected and what does not.
This area is usually defined together with the customer, which means that as a post-editor, you may have to ignore your own “wishes” or “impulses” to correct something. A high degree of professionalism is more important here than perfectionism.
Application scenarios for post-editors
Post-editors can support the productive use of machine translation by actively post-editing the machine-translated text. On the other hand, they can also contribute to the further development of machine translation engines by providing valuable feedback on the quality of the translation to the developers
However, they also want to take on the management of projects that involve machine translation. It therefore makes sense to trust one person with the task of taking care of the continuous optimization of the engines.
This task includes, for example, maintaining training material, collecting feedback from post-editors, internal communication and education on machine translation so that the entire company is at a uniform level of knowledge and awareness (just think of cases where sensitive data was sent to the free cloud and thus became public).
The fact is that human translators cannot possibly handle today’s mass of content to be translated on their own. It is important to identify which translator is suitable for which scenarios – human, machine or both? Human beings are still indispensable, especially for technically demanding, creative and security-related translations. With the help of MT and post-editing, they can integrate a lucrative and hip job into their portfolio.
Therefore we do not ask ourselves: Man or machine? Rather it must be: Man AND machine!
Would you like to learn more about how you can use machine translation productively in your company and which aspects need to be considered? Visit our presentations at Tekom or our stand in hall D07. You are not at Tekom? No problem either! We will be happy to assist you at any time on all language-related topics!
Picture by Green Chameleon. From Unsplash