Technologies to Improve Accuracy of NMT Output and Efficiency of Post-editing

Track: TAUS | TS5 |   Intermediate |
Thursday, June 13, 2019, 9:00am – 9:45am
Held in: Room F5-7
Takeyoshi Nakayama - Human Science Co., Ltd.
Host: Anne-Maj van der Meer

The neural machine translation (NMT) engines are improving day by day, but many engines do not incorporate the features that the professionals need in the localization industry, such as terminologies and style rules. In this session, we will introduce new technologies to add these features on top of any NMT engine, which will improve the accuracy of NMT raw output and the efficiency of post-editing.

Takeaways: Attendees who are not satisfied with the quality of current NMT engines will be able to improve their engines by adding terminology and style features.