Towards Modeling the Style of Translators in Neural Machine Translation

Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 1193–1199 June 6–11, 2021. ©2021 Association for Computational Linguistics 1193 Towards Modeling the Style of Translators in Neural Machine Translation Yue Wang∗and Cuong Hoang and Marcello Federico Amazon AI hoacuong@amazon.com Abstract One key ingredient of neural machine transla- tion is the use of large datasets from different domains and resources (e.g. Europarl, TED talks). These datasets contain documents trans- lated by professional translators using differ- ent but consistent translation styles. Despite that, the model is usually trained in a way that neither explicitly captures the variety of trans- lation styles present in the data nor translates new data in different and controllable styles. In this work, we investigate methods to aug- ment the state-of-the-art Transformer model with translator information that is available in part of the training data. We show that our style-augmented translation models are able to capture the style variations of translators and to generate translations with different styles on new data. Indeed, the generated variations dif- fer significantly, up to +4.5 BLEU score differ- ence. Despite that, human evaluation confirms that the translations are of the same quality. 1 Introduction Translators often translate the original content with provided guidelines for styles.1 However, guide- lines are supposed to be high level and not com- prehensive. Personal stylistic choices are thus wel- come as creative part of the translator’s job, as long as their translation style consistency is en- sured to the task. By contrast, although neural machine translation (NMT) models (Cho et al., 2014; Sutskever et al., 2014) are trained from these human translations (e.g. Europarl, TED Talks), the models do not explicitly learn to capture the rich va- riety of translators’ styles from the data. This limits their capability to creatively translate new data with different and consistent styles as translators do. We believe that modeling the style of translators is an ∗Y. Wang carried out this work during an internship with Amazon AI. 1See https://www.ted.com/participate/translate/guidelines as an example of translation style guidelines. important yet overlooked aspect in NMT. Our con- tribution, to the best of our knowledge, is to fill this gap for the first time. In particular, our work investigates ways to in- tegrate translator information into NMT, with an emphasis on mimicking the translator’s style. Our study uses the TED talk dataset, with four lan- guage pairs with translator annotations. We present and compare a set of different methods of using a discrete translator token to model and control translator-related stylistic variations in translation. Note that using a discrete token is a common ap- proach to model and control not only specific traits in translation such as verbosity, politeness and speaker-related variances (Sennrich et al., 2016a; Michel and Neubig, 2018)) but also other aspects in NMT such as language ids (Johnson et al., 2017; Fan et al., 2020). However, our study is the first to use such a discrete token to model the style …

Content truncated to 500 words