Towards Debiasing Translation Artifacts

TL;DR

This paper introduces a novel method using the INLP algorithm to reduce translation artifacts in cross-lingual NLP, improving task performance by debiasing translationese in latent space.

Contribution

It extends an existing bias-removal technique to specifically target translationese, demonstrating effectiveness at sentence and word levels in improving NLI accuracy.

Findings

Translationese is effectively reduced at sentence and word levels.

Debiasing translationese improves natural language inference accuracy.

First study to debias translationese in latent embedding space.

Abstract

Cross-lingual natural language processing relies on translation, either by humans or machines, at different levels, from translating training data to translating test sets. However, compared to original texts in the same language, translations possess distinct qualities referred to as translationese. Previous research has shown that these translation artifacts influence the performance of a variety of cross-lingual tasks. In this work, we propose a novel approach to reducing translationese by extending an established bias-removal technique. We use the Iterative Null-space Projection (INLP) algorithm, and show by measuring classification accuracy before and after debiasing, that translationese is reduced at both sentence and word level. We evaluate the utility of debiasing translationese on a natural language inference (NLI) task, and show that by reducing this bias, NLI accuracy improves. To the best of our knowledge, this is the first study to debias translationese as represented in latent embedding space.