Using Auto-adaptive Machine Learning for GetYourGuide — a Case Study

Track: Multilingual AI | TA7 |
Friday, October 28, 2016, 2:15pm – 3:15pm
Held in: Fontaine CD
Anne-Cécile Dousson-Lhéritier - Amazon 
Spence Green - Lilt 
Michel Lopez - e2f

Machine translation (MT) systems are traditionally criticized for poor quality output. Yet combining MT with auto-adaptive machine learning enables a new paradigm of “machine assistance.” Such systems learn from the experience, intelligence and insights of human users, improving productivity by working in partnership, making suggestions and improving accuracy over time. In this session we will relate how e2f and Lilt collaborated, using Lilt’s technology to empower e2f’s human translators in order to meet GetYourGuide’s expectation of translating over one million words in six languages in less than two weeks, resulting in a critical customer success.