The main purpose of this study is to evaluate the effectiveness of Machine Translation Post-Editing (MTPE) training for FL students. Our hypothesis was that with specific MTPE training, students will able to detect and correct machine translation mistakes in their FL. Training materials were developed to detect six typical mistakes from Machine Translation (MT) raw output: Accuracy, Word Order, Official Name, Preposition, Omission, and Formal Style. The training materials include three levels of difficulty: Initial - ability to spot a mistake, Intermediate - ability to classify the type of mistake, and Advanced - ability to correct the mistake. A pretest-posttest design with a control group and a trained experimental group was chosen to test the effectiveness of the training programme. In the posttest, the experimental group could identify and correct more mistakes successfully. and in less time than the control group, especially for omission, official name and preposition. Accuracy, formal style, and word order errors were more difficult to correct. Results suggest that specific MTPE training is not only useful to identify and correct MT mistakes but also a way to incorporate a critical view on machine translation in FL classes.
Zhang, H., & Torres-Hostench, O. (2022). Training in machine translation post-editing for foreign language students. Language Learning & Technology, 26(1), 1–17. http://hdl.handle.net/10125/73466
Language Learning & Technology
University of Hawaii National Foreign Language Resource Center Center for Language & Technology (co-sponsored by Center for Open Educational Resources and Language Learning, University of Texas at Austin)
Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License