Are parts of your corporate texts always identical or similar? Are your texts translated from scratch each time? If you answer both questions in the affirmative, this tip is especially valuable for you and your business. When a translation memory is used, every translated sentence is stored in a database. The system recognizes any identical sentences that occur again in a text and suggests the previous translation (100% match). It also recognizes similar sentences that merely need to be adapted (fuzzy matches). The advantage is obvious: The higher the match rate, the less the translation of the sentence will cost. Plus, the use of a translation memory enables a higher degree of consistency in and between translations.
Remember, however, that it is not possible to achieve the same saving potential with every type of text. For example, our blog posts touch on a wide variety of topics and rarely contain two identical sentences, let alone entire text blocks. Accordingly, the number of fuzzy matches and especially of 100% matches is very low. By contrast, the opposite is the case in our technical documentation. Our translation memory enables us to create consistent translations while saving costs.
In the ideal case, both the terminology database and the translation memory are part of a translation management system (TMS). A TMS is the link that connects customers, translation service providers, freelance translators, terminologists, reviewers, and correctors.
Its central components are a customer-specific translation memory, a terminology database, and a translation tool. In addition, a TMS offers translation management and quality assurance features. The open architecture of a TMS also enables the connection of various systems via interfaces, e.g. for content management, product information management, machine translation, or authoring assistance. This, in turn, ensures a continuous process chain throughout which data can be shared seamlessly.