AI-enhanced TM
Uncovering perfect translation matches that were there all along
How can we meet the global demand for brand-consistent multilingual content? With game-changing clarity between machine translation and TM powered by SYSTRAN and XTM. Partially leveraged TM matches are automatically completed through MT to achieve a working match. Millions of incomplete strands are transformed into complete, persuasive multilingual messages.
AI-enhanced TM – the perfect match, and it was there all along
AI-enhanced TM combines cutting-edge Translation Memory (TM) and Neural Machine Translation (NMT) functionality and creates something greater than the sum of its parts. It enables the machine translation engine to work with mid-level fuzzy matches (75% – 84% match) to deliver full matches. Where a sentence has been partially translated previously, and is therefore stored in the TM, the NMT engine fills in the missing words to deliver the best possible match. AI-enhanced TM takes untapped data and converts it into persuasive multilingual communication.
Why AI-enhanced TM, and why now?
Previously, Translation Memory and the Machine Translation Engine worked side-by-side but not in harmony, with the TM mainly used for “training” the MT engine. Now, in a self-sustaining, interactive process, the Neural Machine Translation Engine effectively trains itself on the fly using the translation memory, opening the door to high quality translation at scale.
Where a sentence is already partially translated, the NMT engine can leverage existing matches and fill in the missing words to deliver a full translation. The segment is completed instantly, requiring little or no rework.
These partial matches, amounting to millions of words, have been dangling on a branch within sight of translators and project managers, but tantalizingly just out of reach. Now, with AI-enhanced TM, we can reach and use these assets, saving unprecedented amounts of time and money.