Please use this identifier to cite or link to this item:
https://doi.org/10.21256/zhaw-22677| Publication type: | Conference poster |
| Type of review: | Peer review (abstract) |
| Title: | Neural machine translation in academic contexts |
| Authors: | Delorme Benites, Alice Benites, Fernando |
| et. al: | No |
| DOI: | 10.21256/zhaw-22677 |
| Conference details: | Swiss Text Analytics Conference – SwissText 2021, Online, 14-16 June 2021 |
| Issue Date: | 15-Jun-2021 |
| Publisher / Ed. Institution: | ZHAW Zürcher Hochschule für Angewandte Wissenschaften |
| Language: | English |
| Subjects: | Machine translation; Academic text |
| Subject (DDC): | 410.285: Computational linguistics 418.02: Translating and interpreting |
| Abstract: | English is the main lingua franca for the international scientific dialogue. However, several studies show that a lesser command of English is a factor of exclusion for many non-native researchers. In medicine, English competence robustly predicts chances of publication in leading journals – with a predictive weight that surpasses total financial investment in research. Non-native researchers often turn to free online neural machine translation tools (NMT) to spare the additional temporal and financial costs imposed by journals’ expectation that submissions are “proofread by a native speaker”. However, general NMT cannot account for the specificities of German academic texts: terminology (i.e. German ad hoc compounding, frequent neologisms, popular "German" terms existing parallel to scientific Greco-latin terms), syntax (i.e. presenting constructions, grammatical subjects and rhematized long subjects) and hedging (i.e. German modal verbs) are the three main issues. By fine-tuning a state-of-the-art neural machine translation system on specific academic corpora, we hope to tackle these issues and to compare which errors were corrected and which were not. Such a domain specific approach could be adopted by institutions to train in-house systems and become a standard tool provided by the institution to their researchers like any other intern application. |
| URI: | https://digitalcollection.zhaw.ch/handle/11475/22677 |
| Fulltext version: | Published version |
| License (according to publishing contract): | Licence according to publishing contract |
| Departement: | Applied Linguistics |
| Organisational Unit: | Institute of Translation and Interpreting (IUED) |
| Published as part of the ZHAW project: | Machine translation for academic texts Machine translation literacy for academics |
| Appears in collections: | Publikationen Angewandte Linguistik |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| 2021_Delorme-Benites_Machine-translation-academic-context_SwissText21.png | 1.78 MB | image/png |  View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.