What is data enhancement?
Data enhancement is the process of using human beings to lay the finishing touches on your multilingual content once our Neural Machine Translation (NMT) engines have generated an immediate translation.
- By combining AI-powered NMT with human enhancement, we provide high-quality multilingual content faster and at a lower price point than traditional workflows.
- By virtue of our monolingual editor, we achieve more than mere translations - your content will read as if originally created in the target language.
- On average, our enhancers correct less than 10 words in every 100 – and we’re aiming for 0!
How do we go about enhancing?
- Cultural norms and register – what kind of document is it, and, consequently, what kind of language should be used?
- Usage and word collocations – would a native speaker use such a phrase in this context?
- Terminology and client-specific style guides – part of the enhancement task is to ensure that the target version of the text follows the style guides provided by the client in any given language, as well as the approved terminology.
In addition to grammar and other elements mentioned above, our enhancers make sure to double-check usage and word collocations, ambiguity, cohesion, and much more.
