Advanced search
1 file | 4.77 MB Add to list

Gutenberg goes neural : comparing features of Dutch human translations with raw neural machine translation outputs in a corpus of English literary classics

Rebecca Webster (UGent) , Margot Fonteyne (UGent) , Arda Tezcan (UGent) , Lieve Macken (UGent) and Joke Daems (UGent)
Author
Organization
Project
Abstract
Due to the growing success of neural machine translation (NMT), many have started to question its applicability within the field of literary translation. In order to grasp the possibilities of NMT, we studied the output of the neural machine system of Google Translate (GNMT) and DeepL when applied to four classic novels translated from English into Dutch. The quality of the NMT systems is discussed by focusing on manual annotations, and we also employed various metrics in order to get an insight into lexical richness, local cohesion, syntactic, and stylistic difference. Firstly, we discovered that a large proportion of the translated sentences contained errors. We also observed a lower level of lexical richness and local cohesion in the NMTs compared to the human translations. In addition, NMTs are more likely to follow the syntactic structure of a source sentence, whereas human translations can differ. Lastly, the human translations deviate from the machine translations in style.
Keywords
LT3, literary machine translation, neural machine translation, quality assessment, lexical richness, cohesion, syntactic divergence, Burrows’ delta

Downloads

  • informatics-07-00032-v2.pdf
    • full text (Published version)
    • |
    • open access
    • |
    • PDF
    • |
    • 4.77 MB

Citation

Please use this url to cite or link to this publication:

MLA
Webster, Rebecca, et al. “Gutenberg Goes Neural : Comparing Features of Dutch Human Translations with Raw Neural Machine Translation Outputs in a Corpus of English Literary Classics.” INFORMATICS-BASEL, vol. 7, no. 3, 2020, doi:10.3390/informatics7030032.
APA
Webster, R., Fonteyne, M., Tezcan, A., Macken, L., & Daems, J. (2020). Gutenberg goes neural : comparing features of Dutch human translations with raw neural machine translation outputs in a corpus of English literary classics. INFORMATICS-BASEL, 7(3). https://doi.org/10.3390/informatics7030032
Chicago author-date
Webster, Rebecca, Margot Fonteyne, Arda Tezcan, Lieve Macken, and Joke Daems. 2020. “Gutenberg Goes Neural : Comparing Features of Dutch Human Translations with Raw Neural Machine Translation Outputs in a Corpus of English Literary Classics.” INFORMATICS-BASEL 7 (3). https://doi.org/10.3390/informatics7030032.
Chicago author-date (all authors)
Webster, Rebecca, Margot Fonteyne, Arda Tezcan, Lieve Macken, and Joke Daems. 2020. “Gutenberg Goes Neural : Comparing Features of Dutch Human Translations with Raw Neural Machine Translation Outputs in a Corpus of English Literary Classics.” INFORMATICS-BASEL 7 (3). doi:10.3390/informatics7030032.
Vancouver
1.
Webster R, Fonteyne M, Tezcan A, Macken L, Daems J. Gutenberg goes neural : comparing features of Dutch human translations with raw neural machine translation outputs in a corpus of English literary classics. INFORMATICS-BASEL. 2020;7(3).
IEEE
[1]
R. Webster, M. Fonteyne, A. Tezcan, L. Macken, and J. Daems, “Gutenberg goes neural : comparing features of Dutch human translations with raw neural machine translation outputs in a corpus of English literary classics,” INFORMATICS-BASEL, vol. 7, no. 3, 2020.
@article{8673762,
  abstract     = {Due to the growing success of neural machine translation (NMT), many have started to question its applicability within the field of literary translation. In order to grasp the possibilities of NMT, we studied the output of the neural machine system of Google Translate (GNMT) and DeepL when applied to four classic novels translated from English into Dutch. The quality of the NMT systems is discussed by focusing on manual annotations, and we also employed various metrics in order to get an insight into lexical richness, local cohesion, syntactic, and stylistic difference. Firstly, we discovered that a large proportion of the translated sentences contained errors. We also observed a lower level of lexical richness and local cohesion in the NMTs compared to the human translations. In addition, NMTs are more likely to follow the syntactic structure of a source sentence, whereas human translations can differ. Lastly, the human translations deviate from the machine translations in style.},
  articleno    = {32},
  author       = {Webster, Rebecca and Fonteyne, Margot and Tezcan, Arda and Macken, Lieve and Daems, Joke},
  issn         = {2227-9709},
  journal      = {INFORMATICS-BASEL},
  keywords     = {LT3,literary machine translation,neural machine translation,quality assessment,lexical richness,cohesion,syntactic divergence,Burrows’ delta},
  language     = {eng},
  number       = {3},
  pages        = {21},
  title        = {Gutenberg goes neural : comparing features of Dutch human translations with raw neural machine translation outputs in a corpus of English literary classics},
  url          = {http://dx.doi.org/10.3390/informatics7030032},
  volume       = {7},
  year         = {2020},
}

Altmetric
View in Altmetric