Measuring the effect of conversational aspects on machine translation quality

9Citations
Citations of this article
85Readers
Mendeley users who have this article in their library.

Abstract

Research in statistical machine translation (SMT) is largely driven by formal translation tasks, while translating informal text is much more challenging. In this paper we focus on SMT for the informal genre of dialogues, which has rarely been addressed to date. Concretely, we investigate the effect of dialogue acts, speakers, gender, and text register on SMT quality when translating fictional dialogues. We first create and release a corpus of multilingual movie dialogues annotated with these four dialogue-specific aspects. When measuring translation performance for each of these variables, we find that BLEU fluctuations between their categories are often significantly larger than randomly expected. Following this finding, we hypothesize and show that SMT of fictional dialogues benefits from adaptation towards dialogue acts and registers. Finally, we find that male speakers are harder to translate and use more vulgar language than female speakers, and that vulgarity is often not preserved during translation.

Register to see more suggestions

Mendeley helps you to discover research relevant for your work.

Already have an account?

Cite

CITATION STYLE

APA

Van Der Wees, M., Bisazza, A., & Monz, C. (2016). Measuring the effect of conversational aspects on machine translation quality. In COLING 2016 - 26th International Conference on Computational Linguistics, Proceedings of COLING 2016: Technical Papers (pp. 2571–2581). Association for Computational Linguistics, ACL Anthology.

Readers' Seniority

Tooltip

PhD / Post grad / Masters / Doc 21

62%

Lecturer / Post doc 5

15%

Researcher 5

15%

Professor / Associate Prof. 3

9%

Readers' Discipline

Tooltip

Computer Science 28

67%

Linguistics 10

24%

Engineering 2

5%

Arts and Humanities 2

5%

Save time finding and organizing research with Mendeley

Sign up for free