What is Translation Memory and why should it matter?

I put this presentation together for a client recently who was struggling to understand what Translation Memory is, and how it works to save money, time, and improve consistency. Feel free to comment below.

3 comments

  1. michael says:

    OK, let me be the dissenting voice here (sorry). I never get a very good and concise translation from your free text translation. Very dissatisfied.

  2. michael says:

    I know…you get what you pay for, right?

  3. Hi Michael,

    I’m sorry you are dissatisfied with our free text translation. You will, of course, also be dissatisfied with any free text translation, as they tend to use very common algorythms (Google’s tool and ours are in fact, one and the same, for example).

    That said, machine translation (which is what the free tool uses) isn’t meant to be 100% accurate, but used a reference tool. On average, machine translation is around 60% accurate – which means it’s 40% wrong. In fact many linguists struggle to understand the output of machine translation, which gives you some idea of teh accuracy.

    We talk about both the limits and the frustrations of relying on free text, or machine translation, on this blog quite often – http://blog.appliedlanguage.com/tag/machine-translation/.

    There are so many variables with certain words that machine translation doesn’t account for contextual factors, or phrases and the many idioms that make up modern languages – as they are unique to each language.

    Probably a moot point – but Greg’s presentation isn’t actually about free text (machine) translation, but translation memory, which are two different things.

    If you do need any text translating with a certain level of accuracy, we can happily provide you with a free quote here – if you pay for your translation online, you can get 25% off our ratecard price as standard.

    Hope this helps.

Leave a Reply

Your email address will not be published. Required fields are marked *

*