Large-Scale English-Japanese Simultaneous Interpretation Corpus: Construction and Analyses with Sentence-Aligned Data

Proceedings of the 18th International Conference on Spoken Language Translation, pages 226–235 Bangkok, Thailand (Online), August 5–6, 2021. ©2021 Association for Computational Linguistics 226 Large-Scale English-Japanese Simultaneous Interpretation Corpus: Construction and Analyses with Sentence-Aligned Data Kosuke Doi1 Katsuhito Sudoh1,2 Satoshi Nakamura1,2 1Nara Institute of Science and Technology 2AIP, RIKEN {doi.kosuke.de8, sudoh, s-nakamura}@is.naist.jp Abstract This paper describes the construction of a new large-scale English-Japanese Simultaneous In- terpretation (SI) corpus and presents the re- sults of its analysis. A portion of the cor- pus contains SI data from three interpreters with different amounts of experience. Some of the SI data were manually aligned with the source speeches at the sentence level. Their latency, quality, and word order aspects were compared among the SI data themselves as well as against offline translations. The results showed that (1) interpreters with more experi- ence controlled the latency and quality better, and (2) large latency hurt the SI quality. 1 Introduction Simultaneous interpretation (SI) is a task of trans- lating speech from a source language into a target language in real-time. Unlike consecutive transla- tion, where the translation is done after the speaker pauses, in SI the translation process starts while the speaker is still talking. With recent develop- ments in machine translation and speech process- ing, various studies have been conducted aiming at automatic speech translation (Pino et al., 2020; Wu et al., 2020; Inaguma et al., 2021; Bahar et al., 2021), including SI (Oda et al., 2014; Zheng et al., 2019; Arivazhagan et al., 2019; Zhang et al., 2020; Nguyen et al., 2021), based on speech corpora. Existing speech corpora can be classified into Speech Translation corpora or Simultaneous In- terpretation corpora, as defined by Zhang et al. (2021). Table 1 lists publicly-available SI corpora. Although a large number of Speech Translation corpora have been published, the number of SI cor- pora remains very limited. Both types of corpora are comprised of audio data and their correspond- ing translations, although how the translations are generated is different. For Speech Translation cor- pora, a translation is based on complete audio data Corpora Language Hours Toyama et al. (2004) En↔Jp 182 Paulik and Waibel (2009) En↔Es 217 Shimizu et al. (2014) En↔Jp 22 Zhang et al. (2021) Zh→En 68 Ours En↔Jp 304.5 Table 1: Existing SI corpora and ours or transcripts; for SI corpora, human interpreters actually do SI. SI corpora are useful not only for the construction of automatic SI systems but also for translation studies. To facilitate research in the field of SI, we are constructing a new large-scale English↔Japanese SI corpus1. We recorded the SIs of lectures and press conferences and amassed over 300 hours of such data. Some lectures have SI data generated by three interpreters with different amounts of expe- rience, as in Shimizu et al. (2014), which enables comparisons of SI differences based on experience. In this paper, we describe the construction of a new corpus and present the results of its analysis. Its design follows the framework of Shimizu …

Content truncated to 500 words