Using Translation Techniques to Characterize MT Outputs

Sergi Alvarez-Vidal, Maria Do Campo, Christian Olalla-Soler, Pilar Sánchez-Gijón


Abstract
While current NMT and GPT models improve fluency and context awareness, they struggle with creative texts, where figurative language and stylistic choices are crucial. Current evaluation methods fail to capture these nuances, which requires a more descriptive approach. We propose a taxonomy based on translation techniques to assess machine-generated translations more comprehensively. The pilot study we conducted comparing human machine-produced translations reveals that human translations employ a wider range of techniques, enhancing naturalness and cultural adaptation. NMT and GPT models, even with prompting, tend to simplify content and introduce accuracy errors. Our findings highlight the need for refined frameworks that consider stylistic and contextual accuracy, ultimately bridging the gap between human and machine translation performance.
Anthology ID:
2025.mtsummit-1.47
Volume:
Proceedings of Machine Translation Summit XX: Volume 1
Month:
June
Year:
2025
Address:
Geneva, Switzerland
Editors:
Pierrette Bouillon, Johanna Gerlach, Sabrina Girletti, Lise Volkart, Raphael Rubino, Rico Sennrich, Ana C. Farinha, Marco Gaido, Joke Daems, Dorothy Kenny, Helena Moniz, Sara Szoc
Venue:
MTSummit
SIG:
Publisher:
European Association for Machine Translation
Note:
Pages:
619–627
Language:
URL:
https://preview.aclanthology.org/mtsummit-25-ingestion/2025.mtsummit-1.47/
DOI:
Bibkey:
Cite (ACL):
Sergi Alvarez-Vidal, Maria Do Campo, Christian Olalla-Soler, and Pilar Sánchez-Gijón. 2025. Using Translation Techniques to Characterize MT Outputs. In Proceedings of Machine Translation Summit XX: Volume 1, pages 619–627, Geneva, Switzerland. European Association for Machine Translation.
Cite (Informal):
Using Translation Techniques to Characterize MT Outputs (Alvarez-Vidal et al., MTSummit 2025)
Copy Citation:
PDF:
https://preview.aclanthology.org/mtsummit-25-ingestion/2025.mtsummit-1.47.pdf