CUNI systems for WMT21: Terminology translation Shared Task

TL;DR

This paper presents a system for terminology translation in English-French, leveraging term lemmatization and conditioning on provided terms, achieving high accuracy in term inclusion while maintaining translation quality.

Contribution

The authors introduce a method that incorporates terminology databases into neural translation by lemmatizing terms and training the model to produce correct surface forms.

Findings

Ranked second in Exact Match metric at WMT21

Effective use of lemmatization for term consistency

Maintains overall translation quality

Abstract

This paper describes Charles University submission for Terminology translation Shared Task at WMT21. The objective of this task is to design a system which translates certain terms based on a provided terminology database, while preserving high overall translation quality. We competed in English-French language pair. Our approach is based on providing the desired translations alongside the input sentence and training the model to use these provided terms. We lemmatize the terms both during the training and inference, to allow the model to learn how to produce correct surface forms of the words, when they differ from the forms provided in the terminology database. Our submission ranked second in Exact Match metric which evaluates the ability of the model to produce desired terms in the translation.