Using a translation memory for translation is not the same as using machine translation. But what exactly is the difference and when does it make sense to use one over the other?
For an outsider, this is a very complex area. It can be difficult to understand that you cannot just run a text through a translation tool and get a complete translated text as you can when you use a MT tool such as Google Translate.
What is a translation memory?
A translation memory is a database that is an integral part of a translation software, also called a Computer Assisted Translation tool (CAT tool) and is used by translators and translation agencies.
The translation memory is created as the translations are made. The translation software stores the individual segments (sentences) and the corresponding translation in a memory. This allows reusing previously translated segments in new translations.
A translation memory works by the computer searching the database for terms and complete sentences that you have translated before and suggesting the best match between the source text and the target text. The translator will then either reject or accept the suggestion.
Read more about the use of a translation tool and translation memories in this blog post ‘ What is a CAT tool (translation tool)? And why is it a good idea to use it for your translations?’
Using a translation tool with integrated translation memory offers clear advantages to both the translator and the client. The client will notice that the quality of the final product is better since terminology and language are consistent. Moreover, the memory is created during translation so that it matches the individual client’s tone of voice. Another important point is that translations will gradually become cheaper the more text that can be reused from previous translations, since any reuse and repetitions result in a discount. This is something you will definitely notice.
What is machine translation?
When you use pure machine translation (MT), the machine does all the work. You simply enter your text to get it translated.
The quality of the translation may differ considerably. If the text is straight forward, the language combination is good and the machine is trained, you may get a good result. However, in other cases the result may be almost useless.
A machine translation can help make translation quicker and cheaper, but will always require postediting.
It is therefore important to emphasise that while machine translation is an excellent tool for supporting the translator’s work, it is a matter om finding the best way for the translator and machine to work together to achieve the best and cheapest product.
Machine translation is an area undergoing rapid development; indeed, after Google introduced Neural Machine Translation, quality has improved considerably. Initially, Neural Machine Translation has been introduced for the major language combinations, and more combinations will be added later.
Is confidentiality guaranteed?
When you use the free version of Google Translate, Google will keep the texts you have translated. The texts will not be stored in a secure environment, so there is a risk that they might end up in the wrong hands. If you choose to use Google’s paid API solution, Google guarantees that your texts remain confidential.
So it would make better sense to contact a translation agency that has already invested in a good and secure solution. More and more translation agencies have started integrating machine translation in their services and at AdHoc Translations, we also offer various solutions to our clients.
An agency will also be able to make sure that the machine is trained with your company’s terminology so that it gradually becomes better at translating your texts.
When does it make sense to use machine translation?
When you use machine translation, you must accept that the translation is of a much lower quality. However, is it sometimes OK that quality is not top-notch?
Well, maybe. You should at least ask yourself this question if you are considering using machine translation.
Machine translation is both quicker and cheaper, but as you can see in this article the quality will never reach the same standard as a translation made by a professional native-language translator
On the other hand, with the constantly increasing volume of content on the Internet, it might make sense to prioritise your translations according to quality and quantity to determine when machine translation can be used sensibly and thus add value to your business.
As a general rule, you should use a professional translator for content that require accuracy, that must be creative and if there is a risk of misunderstanding or misinterpretation.
One option is to use machine translation with postediting by a professional translator for content with a high publishing frequency where requirements for quality are not very high. This might be content such as customer support on social media. You can also choose to use machine translation for internal documents, minutes, etc. which are only intended to create understanding. It can also help break down language barriers standing in the way of hiring foreign labour. So machine translation can help create value for your business.
However, you should always consider that when you make texts intended to arouse interest, create commitment and convince your audience, it will require a professional, human translator.