The rise of the Translation Memory Technology
Thirty years ago TRADOS GmbH (now SDL) sets the standard for CAT technology with the first version of its terminology management tool and later complemented it with a translation memory technology. Translation Memory (TM) tools have been the cornerstone of productivity software in the translation market, as a faster way to translate lots of repetitive content. Building your TM resources might take some time, but once you start getting those precious matches, you realize its real value.
Over the years, while TM and terminology have always been at the heart of SDL’s translation productivity software, the product tool-set evolved to what we call today a whole language platform. Functionality to speed up translation processes, such as predictive typing, glossary suggestions, quality assurance tools, is constantly being added, but making our technology open, extensible and interoperable with other industry standards has been a big focus in the last couple of years.
The Content Challenge
Meanwhile, the rise of the internet, with over 2.7 billion users today and the boom of the digital communication channels, we see massive growth in content. This has pushed the demand for localization and the translation technology to a whole new level. Research figures from the Common Sense Advisory indicate the constant growth of the localization services market, while the number of translators has been pretty much flat over the past decade.
So, to keep up with the huge amount of content is a big challenge. Human translation, even when combined with translation memory software is not enough to cope with the volume of translation demand, and technology providers are constantly looking into ways to break down language barriers across the internet.
The Human and Technology Merge
That’s where machine translation comes into play. Although it’s been around for a long time, only in the last few years has statistical machine translation started to deliver much better quality translation output. Combined with human translation to form the phenomenon of post-editing, machine translation today is seen as a major market trend.
It’s good to see that more and more translators are embracing machine translation to post-edit. By just looking into our own SDL Trados Studio user base, we have hundreds of thousands of people registered for our free machine translation service. Of course, machine translation is not always suited to all content types, and intellectual property must be carefully respected when using public internet services. However, machine translation is a valuable productivity tool, taking the tedium out of human translation for many and providing useful instant translation when human translation isn’t practical.
Now it’s all about improving the machine translation output. That’s what SDL is also focusing on too. With the introduction of SDL Language Cloud, translators now have a simplified access, over the cloud to a set of machine translation engines that have been specifically trained for industries including automotive, consumer electronics, travel, IT, and life science. If you are particularly focused on translating content for any of these areas, these engines could be a real time saver for your translation activities. Having a tailored, better quality machine translation output with the right terminology can significantly reduce the time spent on post-editing.
This is just the beginning of a long-term release plan that at some point will include self-training with your own translation data on top of the industry engines. Many other services and products will be offered as part of the SDL Language Cloud platform too, which will provide a host of opportunities for translators and translation providers.
So, the SDL translation technology journey continues…What do you think will be the next trend in the language technology?
This blog was originally featured on Translator Thoughts.