Translation Memory

Maximising previous translation reuse

Ask a question goto
Request information goto
Support goto

Repetitions welcome

Translation services today involve translating extensive documents into multiple languages for simultaneous delivery within tight deadlines at economical rates while maintaining high and consistent standards. Such a task is no longer manageable without the assistance of computer-aided translation tools. But such systems do not replace the human translator; instead, they supplement the translator's knowledge and skills.
Translation memory systems (TMS) have established themselves as the most successful and reliable type, one that is based on the principle of re-using previous translations. If a new document contains sentences that are identical or similar to ones that have been translated before, those sentences are not translated again but copied directly from the previous translation and then amended by the translator where required.
Technical documentation in particular has a tendency to repeat large numbers of words and phrases. That characteristic makes it an obvious step to refer back to previous translations when starting a new one.
 

Compatible components for efficiency and flexibility

A modern TMS offers a broad range of functions. Thus the ideal translation tool will feature
a translation-specific editor with integrated import and export mechanisms for editing documents created in a wide variety of file formats,
a fully integrated terminology management system,
an interactive alignment tool for converting translations not produced with a translation memory system, and
the translation memory (TM) as the core element of the system.
The translation memory stores previously translated segments of text together with their target-language equivalents as language pairs. The TMS compares the source text for a new translation with the stored, previously translated source text. In doing so it applies two fundamental principles:
Automatic translation of text using previous translations from the TM
Proposal of suggested translations (so-called fuzzy matches) for text that is similar to previous source text
The translator can choose whether to accept the proposed translation, edit it, or translate it entirely from scratch. Every modification is added to the TM so that it is constantly updated and continually refined.

Transit XV – in a class of its own

Unlike other TM systems, Transit XV is not fundamentally a database system but rather an efficient file management system that generates its translation memory from segment-by-segment linking of source files to their equivalent translations. Transit is therefore the only system that takes context into account. Each segment contains information about textual setting and project and subject reference which ensures context-related translation.
Transit is also distinguished by other features such as
Interfaces with all commonly used word-processing and DTP systems
Separation and separate storage of content and layout information
Comprehensive support for all aspects of translation project management
Extensive quality assurance checks (data integrity, spelling and terminology)
Open-system architecture based on XML
Easy integration in external systems via API
Exchange of translation memories with other translation memory systems via TMX
Support of more than 160 languages by the use of double-byte Unicode

Transit XV – a TM solution that pays off

Transit
automates recurring translation tasks
reduces administrative work
rationalises DTP work subsequent to translation
and improves cost-efficiency.
A TM solution with an integrated terminology component also guarantees consistency of language style and terminology and forms an essential basis for the corporate wording policies of international businesses and organisations.
^ to top