Safaba’s EMTGlobal™ helps language service providers (LSPs) reduce operational costs, drive faster translation turnaround and increase translation volumes by rapidly closing the MT <> PE gap.
By transforming the traditional language translation process into a highly-automated one, we help increase vendor productivity, reduce translation process overheads and introduce a significant competitive advantage in an otherwise commoditized market.
The process is simple – machine translation produces suggested translations for each source segment (sentence) that is then either accepted, rejected or edited by a professional translator. This is referred to as machine translation post-editing or PEMT. The higher the percentage of accepted and edited segments is, the greater the productivity gains measured and the impact on an LSP’s bottom line.
Reducing post-editing efforts
For post-edited MT (PEMT) to become a valuable alternative, automated translation must produce highly accurate and relevant translations that enable effective post-editing.
Analysis performed by Welocalize on a live, enterprise-scale post-editing project using iOmegaT post-editing productivity analysis tools, found that on average over 40% of EMTGlobal™ translated segments required no post editing with most remaining segments edited for online publication.
EMTGlobal™ encompasses innovations › and drive greater productivity in post-edited MT for enterprise content.
1) Multifaceted Enterprise Language Adaptation – Aligning translation accurately with industry, regulatory and individual brand language
2) Automated Overnight Learning – Ensuring language alignment continually and quickly improves based on post-edited segments
3) Accurate Cross-Language Styling Projection – Producing true-to-source document structure and text format
4) User Defined Translation Settings – Providing real-time control over corporate terminology and vocabulary preferences
5) Segment-Level Usability Estimations – Assigning scores to help assess the potential value of each translated segment for post-editing
Post-edited MT (PEMT) is becoming an increasingly popular approach for commercial content translation. According to Common Sense Advisory, 25% of translated business content today is post-edited MT. In fact, for certain content types such as documentation, FAQ’s and knowledge bases, the survey shows post-edited machine translation is preferred over human translation by translation services buyers.
Together with our world-class language services partners, Safaba has successfully implemented some of the world’s largest post-edited translation projects to-date helping global enterprises propel their globalization strategy forward.