Advanced search
1 file | 3.40 MB Add to list

Identifying the machine translation error types with the greatest impact on post-editing effort

Joke Daems (UGent) , Sonia Vandepitte (UGent) , Robert Hartsuiker (UGent) and Lieve Macken (UGent)
Author
Organization
Abstract
Translation Environment Tools make translators' work easier by providing them with term lists, translation memories and machine translation output. Ideally, such tools automatically predict whether it is more effortful to post-edit than to translate from scratch, and determine whether or not to provide translators with machine translation output. Current machine translation quality estimation systems heavily rely on automatic metrics, even though they do not accurately capture actual post-editing effort. In addition, these systems do not take translator experience into account, even though novices' translation processes are different from those of professional translators. In this paper, we report on the impact of machine translation errors on various types of post-editing effort indicators, for professional translators as well as student translators. We compare the impact of MT quality on a product effort indicator (HTER) with that on various process effort indicators. The translation and post-editing process of student translators and professional translators was logged with a combination of keystroke logging and eye-tracking, and the MT output was analyzed with a fine-grained translation quality assessment approach. We find that most post-editing effort indicators (product as well as process) are influenced by machine translation quality, but that different error types affect different post-editing effort indicators, confirming that a more fine-grained MT quality analysis is needed to correctly estimate actual post-editing effort. Coherence, meaning shifts, and structural issues are shown to be good indicators of post-editing effort. The additional impact of experience on these interactions between MT quality and post-editing effort is smaller than expected.
Keywords
LT3, post-editing, machine translation, translation quality, post-editing effort, effort indicators

Downloads

  • 8554915.pdf
    • full text
    • |
    • open access
    • |
    • PDF
    • |
    • 3.40 MB

Citation

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

MLA
Daems, Joke, et al. “Identifying the Machine Translation Error Types with the Greatest Impact on Post-Editing Effort.” FRONTIERS IN PSYCHOLOGY, vol. 8, 2017, doi:10.3389/fpsyg.2017.01282.
APA
Daems, J., Vandepitte, S., Hartsuiker, R., & Macken, L. (2017). Identifying the machine translation error types with the greatest impact on post-editing effort. FRONTIERS IN PSYCHOLOGY, 8. https://doi.org/10.3389/fpsyg.2017.01282
Chicago author-date
Daems, Joke, Sonia Vandepitte, Robert Hartsuiker, and Lieve Macken. 2017. “Identifying the Machine Translation Error Types with the Greatest Impact on Post-Editing Effort.” FRONTIERS IN PSYCHOLOGY 8. https://doi.org/10.3389/fpsyg.2017.01282.
Chicago author-date (all authors)
Daems, Joke, Sonia Vandepitte, Robert Hartsuiker, and Lieve Macken. 2017. “Identifying the Machine Translation Error Types with the Greatest Impact on Post-Editing Effort.” FRONTIERS IN PSYCHOLOGY 8. doi:10.3389/fpsyg.2017.01282.
Vancouver
1.
Daems J, Vandepitte S, Hartsuiker R, Macken L. Identifying the machine translation error types with the greatest impact on post-editing effort. FRONTIERS IN PSYCHOLOGY. 2017;8.
IEEE
[1]
J. Daems, S. Vandepitte, R. Hartsuiker, and L. Macken, “Identifying the machine translation error types with the greatest impact on post-editing effort,” FRONTIERS IN PSYCHOLOGY, vol. 8, 2017.
@article{8528320,
  abstract     = {{Translation Environment Tools make translators' work easier by providing them with term lists, translation memories and machine translation output. Ideally, such tools automatically predict whether it is more effortful to post-edit than to translate from scratch, and determine whether or not to provide translators with machine translation output. Current machine translation quality estimation systems heavily rely on automatic metrics, even though they do not accurately capture actual post-editing effort. In addition, these systems do not take translator experience into account, even though novices' translation processes are different from those of professional translators. In this paper, we report on the impact of machine translation errors on various types of post-editing effort indicators, for professional translators as well as student translators. We compare the impact of MT quality on a product effort indicator (HTER) with that on various process effort indicators. The translation and post-editing process of student translators and professional translators was logged with a combination of keystroke logging and eye-tracking, and the MT output was analyzed with a fine-grained translation quality assessment approach. We find that most post-editing effort indicators (product as well as process) are influenced by machine translation quality, but that different error types affect different post-editing effort indicators, confirming that a more fine-grained MT quality analysis is needed to correctly estimate actual post-editing effort. Coherence, meaning shifts, and structural issues are shown to be good indicators of post-editing effort. The additional impact of experience on these interactions between MT quality and post-editing effort is smaller than expected.}},
  articleno    = {{1282}},
  author       = {{Daems, Joke and Vandepitte, Sonia and Hartsuiker, Robert and Macken, Lieve}},
  issn         = {{1664-1078}},
  journal      = {{FRONTIERS IN PSYCHOLOGY}},
  keywords     = {{LT3,post-editing,machine translation,translation quality,post-editing effort,effort indicators}},
  language     = {{eng}},
  pages        = {{15}},
  title        = {{Identifying the machine translation error types with the greatest impact on post-editing effort}},
  url          = {{http://dx.doi.org/10.3389/fpsyg.2017.01282}},
  volume       = {{8}},
  year         = {{2017}},
}

Altmetric
View in Altmetric
Web of Science
Times cited: