Wikipedia:Google Translate

In November 2016, Google Neural Machine Translation was introduced,[1] resulting in a significant increase in the accuracy of machine translation. What does this mean for the use of Google Translate on Wikipedia?

Recent research

edit
  • Jackson, Jeffrey L; Kuriyama, Akira; Anton, Andreea; Choi, April; Fournier, Jean-Pascal; Geier, Anne-Kathrin; Jacquerioz, Frederique; Kogan, Dmitry; Scholcoff, Cecilia; Sun, Rao (30 July 2019). "The Accuracy of Google Translate for Abstracting Data From Non–English-Language Trials for Systematic Reviews". Annals of Internal Medicine. 171 (9): 677. doi:10.7326/M19-0891. ISSN 0003-4819.
  • DuBose, Joy (3 October 2019). "Russian, Japanese, and Latin Oh My! Using Technology to Catalog Non-English Language Titles". Cataloging & Classification Quarterly. 57 (7–8): 496–506. doi:10.1080/01639374.2019.1671929.
  • Michael, Erica B.; et al. (2018). "Human use of machine translation to extract information from texts". In Lacruz, Isabel; Jääskeläinen, Riitta (eds.). Innovation and Expansion in Translation Process Research. John Benjamins Publishing Company. pp. 191–. ISBN 978-90-272-6475-6.

Pros and cons

edit
Pros
  • The accuracy of Google Translate continues to improve, and in many cases approaches the accuracy of human translation
  • Use of non-English sources can help counter systemic bias on Wikipedia, which skews to Anglocentric and Eurocentric perspectives
Cons
  • Accuracy may not be sufficient for all uses, and human translation is still more accurate
  • The reliability and due weight of non-English sources must be established
Caveats
  • Translation into English is usually more accurate than translation from English
  • Translation from languages related to English (i.e., Indo-European languages) is more accurate than translation from unrelated languages

References

edit
  1. ^ Lewis-Kraus, Gideon (14 December 2016). "The Great A.I. Awakening". The New York Times. Retrieved 1 February 2020.
edit