Controlling Extra-Textual Attributes about Dialogue Participants -- A Case Study of English-to-Polish Neural Machine Translation
This addresses the problem of translation errors in morphologically rich languages for machine translation systems, particularly in TV dialogue, though it is incremental as it builds on existing attribute control approaches.
The paper tackles the challenge of translating English to Polish, a morphologically rich language, where extra-textual attributes like gender and number must be inferred, by proposing methods to control these attributes using external metadata in TV dialogue translation, achieving an improvement of +5.81 chrF++/+6.03 BLEU with the best model.
Unlike English, morphologically rich languages can reveal characteristics of speakers or their conversational partners, such as gender and number, via pronouns, morphological endings of words and syntax. When translating from English to such languages, a machine translation model needs to opt for a certain interpretation of textual context, which may lead to serious translation errors if extra-textual information is unavailable. We investigate this challenge in the English-to-Polish language direction. We focus on the underresearched problem of utilising external metadata in automatic translation of TV dialogue, proposing a case study where a wide range of approaches for controlling attributes in translation is employed in a multi-attribute scenario. The best model achieves an improvement of +5.81 chrF++/+6.03 BLEU, with other models achieving competitive performance. We additionally contribute a novel attribute-annotated dataset of Polish TV dialogue and a morphological analysis script used to evaluate attribute control in models.