What is Amazon MT?
Amazon MT uses neural machine translation to offer high quality and fast language translation. In this context, neural refers to deep learning models that have been designed to deliver more accurate and decidedly more human-sounding translations than statistical MT.
Two key benefits of Amazon MT are customization and scalability, making it ideal for all projects, whether they’re easy or tricky, big or small.
Amazon MT currently supports 54 languages (full list here).
How does it work with memoQ?
memoQ’s AmazonMT plugin has an especially useful feature called Named Entity Translaton Customization. In short, this means that context is taken into account when translating a word or a phrase.
As many users will be familiar with, machine translations often falter when it comes to context and translate terms that should have been recognized as proper nouns, which in some instances can lead to poor or misleading translations. To resolve this, users can upload or select custom terminologies.
With the release of memoQ 9.3, Amazon MT is at last available. According to Gergely Vándor, Solution Architect at memoQ, this has been one of the most requested integrations and users will be thrilled by its introduction.