The use of translation memory systems has nothing to do with machine translation.
I personally
complete all translations – everything saved in the translation memories (databases)
has been previously translated by me. An individual and customised database is created for
each client as well as for each direction of translation
(e.g. one for English – German and
one for German – English).
When the translation is prepared and the quotation set up, all of your documents for translation are analysed so that the scope of text to be translated is precisely determined. Internal redundancies (repetitions) are reliably identified and the exact number of words of complete as well as partial matches are analysed in the event of follow-up translations. This creates a dual benefit for you: Consistent translation ensures the translation quality and, additionally, matches are discounted.
Prior to and during translation, the source text segments are compared against the database; matches are identified by the translation software and displayed accordingly. Only completely identical translation units already contained in the database are automatically included in the target language document.
Apart from standard functions, such as spell check and punctuation check, comparison with your customer-specific terminology ensures a high quality standard.
I use the following translation tools:
> SDL Trados Studio (current version)
> SDL Trados 2007