2.2 - Source Text Segmentation

Creating a Translation Memory
A TM system is essentially a type of database, it is a piece of software that allows translators to store and retrieve information. Translators may find it more useful to create separate TMs for different subject fields (e.g. in order to reduce false hits caused by homonymy), or to have different TMs for different clients (to respect specific preferences related to terms and style).
(Bowker 2008, 108)
In the majority of cases, the CAT tool allocates a certain folder on a hard drive for storing separate files, in which all previously translated segments are stored in a TM. Depending on the CAT tool of your choice you have a more or less limited range of possibilities for defining rules for splitting text into (translation) segments.
By default, it is the dot sign (".") that triggers carriage returns done automatically by the software at the initial stage of loading a source text into the CAT tool. As was mentioned in previous section, the rules may be in most cases modified by the user.
__________________
Reflection #1: Why is it sometimes good to have more than one translation memory (TM) ?
Sometimes, for instance, when the source is an English text with a lot of high numbers, where dot sign separates thousands, it is even necessary to define other segmentation rules. Similar problem arises when ordinal numbers are used in numbered-off list items.
(Un)Fortunately, (only) the more "grown-up" CAT tools have rules treating segmentation problem automatically, based on the chosen source language. Therefore, for instance, SDL Trados (Figure 1) users may not need to worry about this problem. In OmegaT+ (Figure 2) this feature is not yet in-built, but numerous segmentation rules may be defined with the use of REGULAR EXPRESSIONS contained in the user manual.
Irrespective of the CAT tool the source file is processed and split into segments. Translators are then free to translate each segment. Original source text segments with respective translations are called TRANSLATION UNITS or TUs. As all that is done automatically for them, translators may focus on translation of the content and thus work at a faster rate.
__________________
Reflection #2: What is the difference between a translation unit and translation segment?
For suggested answers to reflection questions click the following button: