Online Learning for Neural Machine Translation Post-editing

TL;DR

This paper explores online learning techniques to improve neural machine translation post-editing, demonstrating significant quality enhancements and effort reduction through a new optimization algorithm and comprehensive comparisons.

Contribution

It introduces a novel online learning algorithm for neural machine translation post-editing and provides a thorough comparison with existing methods.

Findings

Significant improvements in translation quality.

Reduction in post-editing effort.

Effective online learning algorithm for NMT post-editing.

Abstract

Neural machine translation has meant a revolution of the field. Nevertheless, post-editing the outputs of the system is mandatory for tasks requiring high translation quality. Post-editing offers a unique opportunity for improving neural machine translation systems, using online learning techniques and treating the post-edited translations as new, fresh training data. We review classical learning methods and propose a new optimization algorithm. We thoroughly compare online learning algorithms in a post-editing scenario. Results show significant improvements in translation quality and effort reduction.