Machine Translation

Neural Machine Translation (NMT)

Until recently, machine translation was not promising in Turkish language as proven by its performance so far. As one of the agglutinative languages, alignment of the words in Turkish language is reverse when compared to the Indo-European languages like English. This makes translation a tough job!

As a result of Turkey's geography, Islamic background and cultural heritage in between Europe, Russia and Middle East; stretching out on the lands where many ancient civilizations passed away, Turkish translation is a demanding task.

NMT powered by past TMs 

NMT engine can be put into application exclusively for your project. In other words, when NMT is fed by the translation memories generated out of your previous translations, performance will be more editor friendly and compatible with your industrial terminology and style.  

Post-editing

Regarding NMT, everything is developing in a rocket speed. Yet, human post-editing is indispensible. Post-editing of the translations by MT is a new kind of job somewhere between translation and proofreading. Out of this reason, traditional  editors need to be trained and perfectly instructed.

Extra-large content

NMT engine is able to learn from previous TMs. This makes it possible to deal with large amount of content much faster by running NMT engines and human post-editing. 

Project management

If NMT engine is used in any project, this would mean that the project is large enough where several project members are involved. Hence, a project manager must supervise all throught the project stages including glosarry adaptation, post-editing, quality assurance and monitoring.

Recent case studies

Amazon trusted us in our machine translation post editing services.

Thanks to our project managers, ideally specialized post editors will be perfectly identified, booked and assigned to your projects, understanding the requirements of your purpose.