Multi-perspective Alignment for Increasing Naturalness in Neural Machine Translation

TL;DR

This paper introduces a novel reinforcement learning-based method for neural machine translation that enhances naturalness and lexical diversity without sacrificing translation accuracy, evaluated on English-Dutch literary translation.

Contribution

It proposes a multi-perspective alignment approach that balances naturalness and content preservation in NMT, inspired by reinforcement learning from human feedback.

Findings

Translations are lexically richer and more human-like.

No loss in translation accuracy observed.

Reduces machine and human translationese.

Abstract

Neural machine translation (NMT) systems amplify lexical biases present in their training data, leading to artificially impoverished language in output translations. These language-level characteristics render automatic translations different from text originally written in a language and human translations, which hinders their usefulness in for example creating evaluation datasets. Attempts to increase naturalness in NMT can fall short in terms of content preservation, where increased lexical diversity comes at the cost of translation accuracy. Inspired by the reinforcement learning from human feedback framework, we introduce a novel method that rewards both naturalness and content preservation. We experiment with multiple perspectives to produce more natural translations, aiming at reducing machine and human translationese. We evaluate our method on English-to-Dutch literary translation, and find that our best model produces translations that are lexically richer and exhibit more properties of human-written language, without loss in translation accuracy.