Have you noticed how the automatic translation built in to your e-mail client has improved over the last few years? Translation providers noticed too and had to adjust to the new game rules. You might be surprised just how much the AI and neural networks development hit professional translation and reshaped the landscape of services offered to end clients, mainly in the technical field.
Today, operating a flexible translation process would be unthinkable without employing AI-based neural networks and deep learning, translation server platforms, and varying degrees of machine translation to achieve faster turnaround and improve quality while keeping the cost down.
Through our almost 40 years of experience in translation of technical content, we can clearly see the opportunities and also the challenges brought upon the industry by these technological leapfrogs.
And we will tell you a little secret: global translation providers employ commercial machine translation (MT) platforms and text databases in the first round of the translation process. idioma also uses its own Technical Contents Corpus, which is a unique text repository of data resources collected over several decades It significantly improves the quality of raw MT output (unedited by humans). After the source material is pre-translated against this repository as well as by machines, native human translators take over and use the output to speed up the localization process, especially for repetitive content. So old-fashioned human translators still exist, but their role in localizing text content has changed.
For some, read big, languages adding the machine translation to the workflow promises reduced delivery terms so much that translators fully transit into the role of (post-)editors. As editors, translators instead review, improve and sometimes completely rewrite machine-translated text to make it acceptable and comprehensible, and to assure terminology resources are correct and in place for future use. To be a successful professional translator today, linguistic excellence is just a part of the mix. And to run a successful translation business means to invest in new technologies, constantly innovate and teach the machines how to serve us better. Translation, at least in the technical subject field, has become less of an art and more of an analysis and pure data operation.
These circumstances have created an increasing demand for translation output that is faster and cheaper than regular human translation, and where stakeholders knowingly sacrifice human style and format (just think of how weird your GPS sounds sometimes).
Albeit not matching human translation in terms of style, expressive power and format, translation by machines is definitely not recommendable for many subject matters, however machine translation with post-editing (MTPE) can deliver viable results that are in increasing demand mainly in the tech e-commerce sphere and for highly repetitive technical content. In a few years, a brand-new AI-driven service has seen its light in the translation industry – the same industry that not so long ago relied on typewriters and simple word processing software.
If you want to consider MTPE as an alternative for professional human translation, bear in mind that overall quality of machine translation very depending on the language and language combination. While Romance languages, for example, achieve quite impressive and complex results compared to just a few years ago, others, such as Japanese, are still problematic for the machines to process. Currently, idioma offers full-scope translation service into 70+ world languages in various combinations, while 11 languages are now available for MTPE: English, Swedish, Danish, German, Dutch, French, Italian, Spanish, Portuguese, Polish, and Russian. All managed and delivered in compliance with our ISO-certification and DIN ISO-accreditation for multilingual MTPE services.
To learn more, meet us at our stand B 25B at Industrimässorna 2019!