My boss wants me to get our website translated from English into Spanish and one of the agencies I spoke to said they can do it and will give us a translation memory. What’s a translation memory and what’s it for? I need to know quickly please.
Thanks for your question. I’ve explained what you need to know below. Please let us know if you have any further questions.
What are translation memory tools?
Translation memory (TM) tools are software packages designed to improve the efficiency of the translation process by making it easier to deal with chunks of text which are repeated or reoccur in different variations throughout the document.
TM software includes functions for managing terminology, allowing the translator to be consistent in the use of terminology and to store terminology for future projects.
They also include quality assurance features that enable a number of automated checks to be run on translations. For example, it’s possible to check whether numbers have been transposed correctly, whether segments, or parts of sentences, exceed a specified length, or whether incorrect terms, expressions, or spellings have been used.
Although TM tools are great help, they can only be used for certain types of text – more details later in this article.
How do Translation Memory tools work?
TM tools work by storing “segments” of text in a memory as the translator working on your document translates them.
The software checks to see if a segment reoccurs elsewhere in the text. If it does, the translator can display its previous translation from the memory instead of having to re-translate it – but only if the previous translation is appropriate in the current context. This last point is important as it is always the human translator, not the software which makes the final decision.
If a segment is similar but not exactly the same as a previous one, the translator can summon a partial (or “fuzzy”) match from the memory for editing.
What types of texts do Translation Memory tools work well with?
As TM tools are software, they can only work with documents which are in digital form, such as the text of word-processing files.
They work best with documents which contain a large proportion of repeated or partially repeated text. Examples include:
- Software specifications, help files etc.
- E-commerce websites with large catalogues of similar products
- Instruction and operator manuals
TM tools are also useful for some kinds of documents which do not contain repetitions, but are updated periodically based around the same content e.g. monthly reports presenting the same data each month, contracts which are modified periodically.
What are the benefits of using TM tools?
The great thing about TM tools is they help everyone involved in the translation process.
- Cost: You save money by paying lower fees for repetitions and “fuzzy” matches. Typically we charge you 100% of the translation cost for translating the first instance of a segment’s occurrence, 60% for fuzzy matches and only 30% for pure repetitions.
- Time: Documents can be translated and delivered more quickly. Using advanced features of the software documents can even be split up and distributed to teams of translators meaning your job is done in next to no time with no loss of quality.
- Consistency: It is easier for translators to maintain consistency, especially over large documents.
- Portability: You get to keep the translation memory files so you can use different translators in the future with no detriment to consistency.
- Localisation: TM tools can be a very useful tool for aiding the localisation of HTML and other software files
What commercial translation memory packages are there?
Some of the most popular translation memory tools include Trados, DéjàVu and SDLX.