A Case Study on How Neural MT Has Improved the Quality of MT Output

Track: TAUS | TS4 |
Wednesday, February 27, 2019, 4:15pm – 5:00pm
Held in: Kelantan
Michael Cárdenas - Local Concept 
Denis Gachot - SYSTRAN
Host: Anne-Maj van der Meer

In this presentation we will share a case study where translators evaluated a neural machine translation (MT), a hybrid MT and a traditional human-based translation. Translators were also asked whether hybrid and neural MT throughput were good enough to edit, and what the time spent editing all three types of output was.

Takeaways: Attendees will learn how neural MT has improved the grading of the throughput.