Technology never stays static. New ideas and tools come out and they seem revolutionary, amazing. Your work is cut in half, you make more money, and enjoy your work more! Slowly, the dazzle fades, and the new technology or tool becomes the usual.
You get used to it, and begin to resent it just like you resented your work before. You forget your awe when it was first introduced and it simply becomes the same old same old. Then someone comes up with another step forward and you’re excited all over again – for a while.
That’s the story of Translation Memory tools for those of us in the translation business. When TM was first introduced it was controversial, but also exciting: The idea that you could accumulate a database of previously translated material that would speed up your current work and spare you from having to constantly translate common phrases was exciting, and revolutionary. For a while TM was all the rage and everyone talked endlessly about how much better translation work was with it.
Today? TM is old-school, and translation professionals are demanding something new and better. And a company named Lingotek is hoping to provide it.
Search Engine Technology
The limitations of translation memory tools have been well-documented: You are relying on very close matching to get results. Even minor deviations can prevent a match, meaning that you see an awful lot of ‘match not found’ even in work that has a deep database of previous translations to draw on.
Lingotek’s solution is to harness the concepts and algorithms behind search engines. After all, when you fire up Google and type in a search, how often do you get ‘match not found’? Almost never. Lingotek’s idea is to combine the TM concept of a database of collected successful translations with a search engine-like interface that can return much more useful results rather than simply exact or very close matches. In fact, the company claims they can increase the speed of translation work by 800% and allow translators to double their money.
Add In the Internet
Additionally, Lingotek is integrating the Internet and its vast resources of data directly into their tools, so that you’re not merely searching your own database of translation memory, but many databases. This is an idea that’s been a long time coming, as many translation tools were developed in the 1990s before the Internet became the everyday tool it is today, and as a result many translation tools are curiously Internet-blind.
On the other hand, the Internet aspect makes me nervous. I know I can trust my own translation memory databases, because I know I am a good translator. What do I know about the other data sources a tool like Lingotek’s might bring in? It makes me nervous to rely on other people’s translation memory, frankly.
Still, any tool that could potentially double my productivity is worth exploring, and this may be the next big thing.
Image courtesy institutcaraluna.com