Oversættelseshukommelser

Oversættelseshukommelser, often referred to as Translation Memory (TM), are specialized databases used in the field of translation to store previously translated segments of text. These segments are typically sentences or phrases. When a translator works on a new document, the translation memory software scans the source text and compares it against the existing entries in the database. If an exact or near-exact match is found, the software presents the stored translation to the translator. This process significantly speeds up the translation workflow and ensures consistency across different projects, especially for recurring terminology and phrases.

The core principle behind translation memory technology is to leverage past translation efforts. Each time a translator accepts a TM suggestion, or manually creates a new translation pair, it is stored in the database for future use. This creates a growing repository of translated content specific to a client, subject matter, or language pair. Translation memories are instrumental in reducing translation costs and turnaround times, as translators do not need to re-translate identical or similar content. They are a cornerstone of modern computer-assisted translation (CAT) tools, working alongside other features like terminology management. While TM provides significant benefits, human review remains crucial to ensure the accuracy, nuance, and cultural appropriateness of the final translation.