Everybody knows that there are three main drivers when it comes to a translation project: time, cost, and quality. These drivers counter and offset each other: if a project is rushed and has to be done in a short amount of time, costs will be higher and the quality will be affected. If the project has a very small budget, some quality must be sacrificed. If the translation must be of very high quality, costs will be higher and more time will be needed.
Of course, everyone wants their translations to be done as fast as possible, at a high level of quality, at the lowest cost possible. Some trade-offs must be made; however, there’s a new phenomenon sweeping the translation industry that’s making it possible to lower your costs while keeping quality high. It’s called post-editing, and it’s made possible by an amazing development called machine translation. Allow me to take a step BACK IN TIME, in order to provide this discussion with some history and context.
In the late 1940s, as the Computer Age was just beginning, discussions began taking place regarding the possibility of using computers as tools to translate languages. Theories were formed and proposals were made, and on January 7, 1954, the first demonstration of machine translation (MT) was performed, through a collaboration between IBM and Georgetown University. It was a glorious event, and it paved the way for extensive funding and development of MT throughout the United States and the world.
Fast forward to 2017, and MT has come a long way. Sophisticated systems and algorithms have been developed to make machine translation more accurate than ever; beyond this, a space-age new technology called neural machine translation has been developed, which uses artificial neural networks based on the biological neural networks of our own brains, and is capable of learning on its own.
All this brings us to the advent of post-editing. As much as MT has developed and advanced since 1954, it still can’t be trusted (yet) to capture the nuances of language the way a native human speaker can. Post-editing is a technique that was developed that allows you to use MT without losing the human touch. It’s a matter of simply using MT to translate a text, and then having a human quickly review the translation before passing it on to the editing stage. Time and costs stay low, and quality remains high. A win-win!
Ask a representative about our services today.
Sources:
https://www.hutchinsweb.me.uk/Nutshell-2005.pdf
