Adaptive machine translation

Machine translation with online learning


An adaptive machine translation system learns from human feedback and adapts its output on the fly. Adaptive machine translation is applicable to post-editing workflows.

In adaptive machine translation, the system is customised while the human post-editor fixes the machine translation output, instead of after batch retraining.

Adaptive machine translation is an example of online machine learning and human-in-the-loop (HITL).

API support

Adaptive machine translation first became practical with the rise of neural machine translation.

Lilt launched its adaptive machine translation feature in 2017.

ModernMT launched its adaptive feature in 2019. Amazon Translate launched Active Custom Translation in 2020. SYSTRAN launched Neural Fuzzy Adaptation in 2021. Language Weaver launched its adaptive feature in 2022.

Workflow

  1. The system generates a machine translated output.
  2. The output is fixed or approved.
  3. The system learns from the correction.
  4. The system reduces the chance of repeating the error.

Adaptive machine translation reduces the time a human takes to edit a machine translation suggestion.


Want to learn more about Adaptive machine translation?


Edit this article →

Machine Translate is created and edited by contributors like you!

Learn more about contributing →

Licensed under CC-BY-SA-4.0.

Cite this article →