Safaba Machine Translation support was added to Memsource Cloud. The integration enables users to access trained Safaba MT engines inside Memsource project management interface in real time. Support for post-editing analysis is provided.
To use, add a customized Safaba engine in Settings – Machine Translation, and then select it in Create New Project window.
With these options selected, you can always pre-translate files with your trained engines – to use “as is” or to post-edit after.
Technology that Safaba develops allows training machine translation engines. It adds company-specific and domain specific content to baseline engine. Terminology and frequently used phrases can be translated correctly, resulting in better quality MT and reduced costs for translation and post-editing.
Safaba’s clients include language service providers and global enterprises, such as Paypal and DELL. With DELL, content is translated daily into nearly thirty (30) different languages by Safaba’s automated translation solutions in collaboration with Welocalize.
With Safaba technology, globalization teams and customer support organizations can dramatically reduce globalization costs associated with translation; increase translation volumes while reducing time-to-market; and deliver multilingual content at more customer experience touch points. Analysis performed by Welocalize on an enterprise-scale post-editing project found that on average over 40% of translated segments required no editing with a number of target languages achieving 60% ‘Perfect’ translations.