Evaluating Domain-Adaptive Machine Translation

Choosing the right Machine Translation for a project is hard: the performance of every engine varies across domains and language pairs, and changes often. With domain-adaptive NMT it's even harder, as there's an additional difference in training corpus requirements and complex cost of ownership models.

In this talk, we describe the approach to address those issues. A large-scale reference-based scoring filters candidate MT engines and identifies a "decisive" set of segments for human LQA. A specific LQA procedure is tuned to choosing the best rather than just an acceptance. Also, we look at the learning gradient to estimate if the same engine will be suitable as more data become available.

About the speaker:

About the speaker: Konstantin Savenkov is CEO of Intento, Inc. After earning a PhD in 2008, he led research and development efforts for online content services, then worked as CTO at Zvooq and as a chief operating officer at Bookmate. In 2016, he contributed his experience in artificial intelligence (AI), tech and operations to launch Intento, Inc., where they build tools to discover, evaluate and use machine translation and other cognitive AI services.

::

Networking 6:30 - 7 PM, and after the talk as time allows
Directions and parking: http://www.imug.org/adobe/
Video: will be announced if available after speaker review
Webcast: RSVP, then contact IMUG via Meetup (*not* in the comments, please): http://www.meetup.com/IMUG-Silicon-Valley/message/?recipientId=11534644

::

Admission is free for IMUG members and Adobe employees, and $5 for all others. IMUG membership is only $20 for the first year, $15 for annual renewal, or $100 for lifetime membership. Click here to join, renew or pay a single non-member event fee via PayPal. Cash and checks also accepted at our events. http://imug.org/join

Please RSVP via Meetup by 4:30 PM two days before the event to help IMUG and our hosts prepare badges for you in advance. After that time it's still OK to RSVP right up to the last minute, as that will help us ensure enough seats for everyone! http://meetup.com/IMUG-Silicon-Valley/

Adobe® Connect™ webcast: If you can't join us in person, RSVP and then contact us via: http://www.meetup.com/IMUG-Silicon-Valley/message/?recipientId=11534644

Many thanks to Dr. Ken Lunde and Adobe Globalization for hosting IMUG! For maps, detailed directions, and more, see http://www.imug.org/adobe/

::

The hashtag for IMUG events is #IMUG408 and you can follow IMUG on Twitter: http://twitter.com/i18n_mug