At Link Translation, we rely on databases. All data – translation memory and terminology entries, source and target texts, and project data – can be stored centrally on our server. Depending on the settings our clients configure, we may be able to store translated sentences and terminology in their own databases. Attributes such as the subject, relation, and project can be stored along with all language data, enabling our team to find whatever we are looking for with the help of filter settings.

Our system stores preferred terms as well as synonyms, definitions and usage information. By determining misnomers, we can stop words that are to be avoided for any reason from creeping into your texts. The translation editor shows us suitable terminology entries for the sentence that is currently being edited. In the translation memory, we can easily check in which context a word was used in the past.

In the translation editor, all relevant information is displayed in a clearly structured form to our translators. Whether we are translating a Word or XML file or localizing software, we always work in the same environment. Apart from the source text and the target text, our translators can see search hits from the translation memory and the terminology database. With just one click, a preview can be generated for most formats to check layout.

Every time we set up a project in our system, we are asked to select the source texts, target languages, due dates and quality criteria. During check-in, source documents are split by format and content and analyzed. Matches from the translation memory can also be inserted in the target text as pre-translation. Throughout the project period, we can monitor the progress of individual subtasks. Once a project is finished, a report showing the workload, repetition rates and quality can be used as the basis for control and billing.

Our system works with numerous quality assurance functions. Certain error sources are analyzed and displayed during translation. Untranslated paragraphs and ignored translation memory entries are also analyzed automatically. The report at the end of a project can systematically present error types and sources, enabling the optimization of future projects.