Analysing Coreference in Transformer Outputs

Abstract

We analyse coreference phenomena in three neural machine translation systems trained with different data settings with or without access to explicit intra-and cross-sentential anaphoric information. We compare system performance on two different genres: news and TED talks. To do this, we manually annotate (the possibly incorrect) coreference chains in the MT outputs and evaluate the coreference chain translations. We define an error typology that aims to go further than pronoun translation adequacy and includes types such as incorrect word selection or missing words. The features of coreference chains in automatic translations are also compared to those of the source texts and human translations. The analysis shows stronger potential translationese effects in machine translated outputs than in human translations.

Bridging Background Knowledge Gaps in Translation with Automatic Explicitation

“…Explicitation in Contemporary Works. Many of contemporary works with the term "explicitation" focus on discourse MT (Hoek et al, 2015;Webber et al, 2015) and mainly developed by Lapshinova-Koltunski et al (2019, 2020. However, despite the broad coverage of the term, the explicitation in these studies are limited to insertion of connectives or annotations of coreference in the target side.…”

Contrastive Learning for Context-aware Neural Machine TranslationUsing Coreference Information

“…In practice, context-aware models do not leverage target-side contexts struggle to maintain these kinds of coreference consistency Lapshinova-Koltunski et al, 2019) because of the asymmetric nature of grammatical components and data distributions. Results show that CorefCL can complement the limitation of sourceonly context-aware models.…”