Translation Memory: How it Works and Why You Need It

Translation Memory: How it Works and Why You Need It

In life, win-win scenarios can be hard to come by. But for translation professionals and their clients, the use of translation memory is something worth celebrating. Translation memory helps translators work faster and more efficiently. Not only does this improve the work flow process for the translator, but it helps ensure the client receives high quality and accurate work in a more timely fashion. And last but certainly not least, the client can save money thanks to this productivity boosting tool. The benefits are pretty indisputable. So let’s take a deep dive into the benefits of translation memory. 

What is a translation memory?

In short, a translation memory acts as a database that can store text that was previously translated. This can include words, sentences, and even whole paragraphs of text. The translation memory breaks up these chosen areas of text into two parts. The source and the target. The source refers to the original language of the text entered into the translation memory. The target is the translation of the original text. Together, these pairs are known as “translation units” or “TUs”. It is possible to have a translation memory for each language pair. Companies who have content translated into multiple languages would need a translation memory for each pair, such as one for German, French, or Spanish

How does it work?

Translation memories are a part of a larger system called a CAT (computer assisted translation) tool. This somewhat broad term generally describes software that professional translators use to increase productivity, catch mistakes, and improve consistency. CAT tools segment text that need translation and in return present the text in a way that is much simpler to translate. 

Translation memories are an important part of CAT tools. When using such a tool, a translator will start by opening the document that needs translation, also known as the source file. Then the translation memory will check to see if any of the recently uploaded content was already translated. The tool will search for any “100% matches” (identical matches) as well as any “fuzzy matches” (similar matches, but not exact). A fuzzy match occurs if there is no exact match for a term or phrase, but there are segments in the translation memory that are similar to the text that a translator is currently working on. Sometimes the difference could be as small as a single word.  

When working through a source file, translators can either accept, modify or override suggestions from the translation memory.  Human judgement is of the utmost importance. Even if a translation memory finds a translation match, the translator has to be very careful and verify the accuracy in a given context. They may have to update the translation accordingly. When a segment doesn’t have a match (this occurs if it hasn’t been translated previously), the translation memory will store the new TUs and this allows the translation memory to continue to grow and improve. 

The Benefits

Utilizing translation memories helps support the localization process. It can also help improve the quality and consistency of a translation job, as well as help complete the work faster. Working with translation memories is not the same as using machine translation (MT) tools. Unlike MT, a tool which automatically translates without human input, professional translators create translation memories. That personal touch accounts for subtleties and contextual scenarios that occur in varying languages and cultures. Both the client and translator will gain confidence and peace of mind by using translation memories throughout their translation work. This tool helps improve the consistency, quality, and speed of work. See, a win-win for everyone.

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

You may use these HTML tags and attributes:

<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>