The translation of idiomatic expressions such as “under the weather” is a current problem in Neural Machine Translation. NMT models incorrectly translate idiomatic expressions’ meanings literally or word-by-word. 

Translation of Idiomatic Expressions 


Baziotis, C., Mathur, P., & Hasler, E. (2023). Automatic Evaluation and Analysis of Idioms in Neural Machine Translation. Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics. https://doi.org/10.18653/v1/2023.eacl-main.267

Automatic Evaluation and Analysis of Idioms in Neural Machine Translation

Neural machine translation systems often mistranslate non-compositional figurative expressions, such as idioms, because their intended meanings cannot be derived by combining the literal meanings of their component words.

Google

University of California, Berkeley

NMT is powered by deep learning and enjoys two advantages over traditional approaches such as rule-based machine translation and statistical machine translation - it does not require linguistic professional input, and it can take into account contextual information in the entire sentence, which leads to high quality and fluent translations. The limitation of NMT is that it requires a high scale of parallel data for model training. 

Neural Machine Translation (NMT)

Low-resource neural machine translation refers to NMT tasks in the setting of having a lack of parallel sentences to train neural machine translation models - i.e. there is not a very large corpus of parallel data between the source and target languages.


Low-resource neural machine translation

Statistical Machine Translation (SMT) and Neural Machine Translation (NMT) are two key paradigms in machine translation:

* **Statistical Machine Translation (SMT):** A phrase-based statistical approach that divides input sentences into word or phrase units and maps them to the most statistically similar translations stored in advance.
* **Neural Machine Translation (NMT):** An end-to-end deep learning approach that applies a single large artificial neural network to translate entire sentences. NMT typically consumes less memory, is faster during inference, and captures contextual information more effectively than the multi-stage pipeline of SMT.

Learn Before

Related

Learn After