If you’re having a lot of translation done then using Translation Memory can make savings on the cost, improve quality of your translations and give quicker turn-around of projects. In all the documents you have, including web pages, presentations, reports, plans, strategies etc., there will be an amount of text which is phrased in exactly the same way.

Our translation memory software stores every piece of text we translate for you. The memory consists of text segments in a source language and their translations into one or more target languages. These segments can be blocks, paragraphs, sentences, or phrases. If the same or similar text appears in another document, the software alerts the translator, who can then insert pre-existing translations from the memory into your documents, review in-context and edit if necessary.

How Does Translation Memory Work?

For any large or ongoing project, with a compatible file type, we use Translation Memory. Every sentence we translate for you will be stored in a database for future use. This brings the benefits of:

• Consistency of translation
  The same terminology and descriptions are used in all your translated material.
• A reduction in turnaround time
  Translators don’t need to re-translate repeated sentences or paragraphs.
• Dramatic cost saving
  Repeated text is only charged at a third of the normal price. This charge is in place as the translator will still check the context of repeated sentences.

The cost savings depend on the amount of new text that matches previously translated text already in the memory. The software analyses the text against the memory and allocates a percentage to each segment.

• A segment which matches exactly is called a “100% match”.
• This level of match also refers to “repetitions” within the text which weren’t previously in the translation memory.
• If a 100% match is preceded and followed by 100% match segments it is referred to as an “ICE (In-Context Exact) match”.
• A level of match between 75 – 99% is referred to as a “fuzzy match”
• 75% match and below is counted as “new text”.

This means we can pass on the discounts to you

• In-Context Exact Matches (ICE) – Free of Charge
• 100% Matches and Repetitions – 65% discount
• ‘Fuzzy’ Matches – 35% Discount

Translation Memory built on-the-fly

Most translation agencies use this type of software which creates and updates memories only after translations are completed, meaning that the benefits are only realised once a translation project is completed.

Global Lingo’s software on the other hand gives you instant savings because it builds itself as our linguists work on your translation. Our software analyses the whole document and makes any fuzzy matches available to the translator, meaning that you’ll receive cost benefits from the very first project.

As your Translation Memory is built up over time the matches increase, meaning that it saves you even more money and time as well as increasing accuracy.

Existing Translation Memories

If you already have existing Memories, that’s great; we can use those, as our system allows us to import them directly and instantly pass on the benefits to you. If you want to use existing memories, make sure you let our Sales Team know before you place an order, so we can ensure we take it into account.

Each Translation Memory is unique to a client because they all have unique ways of communication so we never share Translation Memories across our clients.

We can handle virtually any file format

The real power of our system is that it is primarily XML based, so it converts all non-XML file formats to XML prior to processing and then back into the native format once translated. If the file format is already XML, such as In Design, XHTML, OOXML, ODF or DITA, then our system processes the files without changing them. Having one internal format means that processing is greatly simplified.

We can handle the following file formats:

• XML
• XHTML
• HTML
• ODF (Open Office)
• DITA
• PDF
• Adobe In Design
• Adobe Frame Maker
• OOXML (Microsoft Office 2007)
• Rtf
• Microsoft Office 95 – 2003
• Java Property files
• Net resource files
• Plain text files

Translation Memory is not Machine Translation!

All our translators are highly qualified experienced professionals who use Translation Memory as an aid to their work the software does not replace translators. All texts are thoroughly checked for consistency and accuracy. Machine Translation simply abandons the text to a computer program which can’t possibly understand context and the flow of the communication. The quality of Machine Translation, e.g. Google Translate, is simply not good enough and we will never use that for our clients’ translations.