Trados Basic Concepts |

In this article we will explain some basic terms and concepts to start using SDL Trados.

Translator’s Workbench
This component is part of Trados suite and it is a database system that uses the translation memory concept, a method for capturing, storing and reusing translation. The translations stored are saved in translation memory databases. These linguistic data are stored as bilingual segment pairs (sentence) and each of these bilingual segment pairs is called a “translation unit”.

Every time we create a new translation memory in Translator’s Workbench, it is empty. When we use this translation memory for a translation job, the new translation units that are stored in this database increase the number of units available for reuse.

Translation Unit
When we translate with Translator’s Workbench, this component creates a database where all sentences or segments in their original language and their corresponding linguistic equivalents are stored. These stored segments are known as “translation units”. These translation units that now are stored in the database allow us to carry out very quick and efficient fuzzy matches searches.

Creating a Translation Memory
As specified before, when we create a translation memory in Translator’s Worbench, such memory is empty, but it grows as we work in new translation projects or by importing aligned translation units or aligning projects.

With the first method, interactive translation, each and every time we translate a segment, this translation unit is added to our translation memory. If we find a similar or identical text, Translator’s Workbench will show us our previous translation. Now, you can accept the suggested translation or we can edit it and save a new version of the same. This new version will overwrite the previous one or it will be added as a new unit according to Translator’s Workbench settings.

We can also populate a new or existing translation memory by importing previously translated documents. The Import function allows us to add data exported from other translation memories or from projects aligned using WinAlign. The steps to align documents have been explained in our article Basic Guide on How to Start Working with WinAlign.

Translation memory and Fuzzy Matching Searches
While we are translating, Translator’s Workbench searches among the information stored in the translation memory database and suggests previous translations to use in the new text. The similarity between the new original segment we are translating and the segments already present in the translation memory is the basis for these searches. If there are similarities between the new and existing segments, Translator’s Workbench will suggest them and it will specify its similarity percentage.

In 100% match units there is an exact coincidence between the new segment and the previous segment(s). However, Translator’s Workbench also has the capacity to suggest segments that are not identical, but similar, and this is known as “fuzzy match”. This value can vary from 99% to 30%, but the recommended setting is a value of at least 75%, since a lower value is practically of no help. Translator’s Workbench suggests each fuzzy match in order of importance and highlights the differences between the new and previous segment so that we can choose the translation that is more suitable for the new context.

María Carolina Otero
www.quillslanguage.com

Trados Beginners Guide

Translating PDF files

Be Sociable, Share!