Abstract
Machine translation (MT) systems, especially when designed for an industrial setting, are trained with general parallel data derived from the Web. Thus, their style is typically driven by word/structure distribution coming from the average of many domains. In contrast, MT customers want translations to be specialized to their domain, for which they are typically able to provide text samples. We describe an approach for customizing MT systems on specific domains by selecting data similar to the target customer data to train neural translation models. We build document classifiers using monolingual target data, e.g., provided by the customers to select parallel training data from Web crawled data. Finally, we train MT models on our automatically selected data, obtaining a system specialized to the target domain. We tested our approach on the benchmark from WMT-18 Translation Task for News domains enabling comparisons with state-of-the-art MT systems. The results show that our models outperform the top systems while using less data and smaller models.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
Change history
10 June 2021
The original version of this chapter the table 3 was not correct. This has been corrected.
Notes
- 1.
As of May 2020, Google Translate provided riunione condominiale, which, although correct, is a bit too formal term for this kind of meeting.
- 2.
- 3.
References
Ahmed, F., Shafiq, M.Z., Liu, A.X.: The internet is for porn: measurement and analysis of online adult traffic. ICDCS 2016, 88–97 (2016)
Axelrod, A., He, X., Gao, J.: Domain adaptation via pseudo in-domain data selection. EMNLP 2011, 355–362 (2011)
Bahdanau, D., Cho, K., Bengio, Y.: Neural machine translation by jointly learning to align and translate. In: ICLR 2015 (2015)
Bañón, M., et al.: ParaCrawl: web-scale acquisition of parallel corpora. ACL 2020, 4555–4567 (2020)
Biesinger, R.: Is your software racist? Politico (2018). https://www.politico.com/agenda/story/2018/02/07/algorithmic-bias-software-recommendations-000631
Bojar, O., et al.: Findings of the 2018 Conference on Machine Translation (WMT 2018), Belgium, Brussels, pp. 272–307 (2018)
Buck, C., Koehn, P.: Quick and reliable document alignment via TF/IDF-weighted cosine distance. In: WMT 2016, Berlin, Germany, pp. 672–678 (2016)
Chen, B., Huang, F.: Semi-supervised convolutional networks for translation adaptation with tiny amount of in-domain data. In: Proceedings of the 20th SIGNLL Conference on Computational Natural Language Learning. Association for Computational Linguistics, Berlin (2016)
Dinu, G., Mathur, P., Federico, M., Al-Onaizan, Y.: Training neural machine translation to apply terminology constraints. In: ACL 2019, Florence, Italy, pp. 3063–3068 (2019)
Gao, J., Goodman, J., Li, M., Lee, K.F.: Toward a unified approach to statistical language modeling for chinese. In: ACM TALIP (2002)
Hieber, F., et al.: Sockeye: a toolkit for neural machine translation. CoRR (2017)
Junczys-Dowmunt, M., et al.: Marian: fast neural machine translation in C++. CoRR (2018)
Liu, L., Hong, Y., Liu, H., Wang, X., Yao, J.: Effective selection of translation model training data. In: ACL 2014, Baltimore, Maryland (2014)
Luong, T., Pham, H., Manning, C.D.: Effective approaches to attention-based neural machine translation. In: EMNLP 2015, Lisbon, Portugal (2015)
McCulloch, G.: Covid-19 is history’s biggest translation challenge. wired.com (2020). https://www.wired.com/story/covid-language-translation-problem/
Moore, R.C., Lewis, W.: Intelligent selection of language model training data. In: ACL 2010, Uppsala, Sweden, pp. 220–224 (2010)
Post, M.: A call for clarity in reporting BLEU scores. In: WMT 2018 (2018)
Smith, J.R., Saint-Amand, H., Plamada, M., Koehn, P., Callison-Burch, C., Lopez, A.: Dirt cheap web-scale parallel text from the common crawl. In: ACL 2013, Sofia, Bulgaria, pp. 1374–1383 (2013)
Uszkoreit, J., Ponte, J., Popat, A., Dubiner, M.: Large scale parallel document mining for machine translation. In: COLING 2010, Beijing, China (2010)
Vaswani, A., et al..: Attention is all you need. In: NIPS 2017, pp. 5998–6008 (2017)
Vu, T., Moschitti, A.: CDA: a cost efficient content-based multilingual web document aligner. In: EACL 2021 (2021)
Yasuda, K., Zhang, R., Yamamoto, H., Sumita, E.: Method of selecting training data to build a compact and efficient translation model. In: IJCNLP 2008 (2008)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2021 Springer Nature Switzerland AG
About this paper
Cite this paper
Vu, T., Moschitti, A. (2021). Machine Translation Customization via Automatic Training Data Selection from the Web. In: Hiemstra, D., Moens, MF., Mothe, J., Perego, R., Potthast, M., Sebastiani, F. (eds) Advances in Information Retrieval. ECIR 2021. Lecture Notes in Computer Science(), vol 12656. Springer, Cham. https://doi.org/10.1007/978-3-030-72113-8_44
Download citation
DOI: https://doi.org/10.1007/978-3-030-72113-8_44
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-72112-1
Online ISBN: 978-3-030-72113-8
eBook Packages: Computer ScienceComputer Science (R0)