Unsupervised Neural Machine Translation for Similar and Distant Language Pairs

Abstract

Unsupervised neural machine translation (UNMT) has achieved remarkable results for several language pairs, such as French–English and German–English. Most previous studies have focused on modeling UNMT systems; few studies have investigated the effect of UNMT on specific languages. In this article, we first empirically investigate UNMT for four diverse language pairs (French/German/Chinese/Japanese–English). We confirm that the performance of UNMT in translation tasks for similar language pairs (French/German–English) is dramatically better than for distant language pairs (Chinese/Japanese–English). We empirically show that the lack of shared words and different word orderings are the main reasons that lead UNMT to underperform in Chinese/Japanese–English. Based on these findings, we propose several methods, including artificial shared words and pre-ordering, to improve the performance of UNMT for distant language pairs. Moreover, we propose a simple general method to improve translation performance for all these four language pairs. The existing UNMT model can generate a translation of a reasonable quality after a few training epochs owing to a denoising mechanism and shared latent representations. However, learning shared latent representations restricts the performance of translation in both directions, particularly for distant language pairs, while denoising dramatically delays convergence by continuously modifying the training data. To avoid these problems, we propose a simple, yet effective and efficient, approach that (like UNMT) relies solely on monolingual corpora: pseudo-data-based unsupervised neural machine translation. Experimental results for these four language pairs show that our proposed methods significantly outperform UNMT baselines.

Unsupervised English Intelligent Machine Translation in Wireless Network Environment

“…erefore, both in academic research and industrial applications, machine translation is regarded as a key research task in natural language processing. With the advent of the era of artificial intelligence revolution, machine learning has undoubtedly injected new vitality into machine translation technology [11][12][13][14][15].…”

Enabling Unsupervised Neural Machine Translation with Word-level Visual Representations

“…Recent advancements (Zhang et al, 2017;Artetxe et al, 2018;Lample et al, 2018a;Conneau and Lample, 2019;Song et al, 2019;Han et al, 2021) in UNMT have achieved remarkable milestones. However, there are still certain issues associated with solely utilizing monolingual data for translation, such as domain mismatch problem (Marchisio et al, 2020;Kim et al, 2020), poor performance on distance language pairs (Kim et al, 2020;Chronopoulou et al, 2020;Sun et al, 2021), translationese (He et al, 2022), and lexical confusion (Bapna et al, 2022;Jones et al, 2023). Our work aims to address the issue of lexical confusion by utilizing wordlevel image data, thereby eliminating the need for costly bilingual dictionary annotations.…”

Exploration of the Problems and Solutions Based on the Translation of Computer Software into Japanese Language

“…is approach is based on a large-scale collection of bilingual corpora that are translated into each other. e method is further divided into two types: the statistical-based translation method (SBMT) and the instance-based translation method (EBMT) [11]. e statistical-based translation method (SBMT) was first proposed by Weaver in 1949, but statistical machine translation did not form a systematic theoretical framework in the following decades.…”