What is it about?
Most companies use their specialized corporate language in the documents that they need to translate. The paper shows that a specialized translator performs significantly better than any general translator. Statistical translation might still slightly outperform neural translation in terms of BLEU metrics. Human evaluation, however, shows supremacy of the neural method.
Featured Image
Photo by Franck V. on Unsplash
Why is it important?
The paper gives a number of translation examples that may help reader to see the cons nad prons of statistical and neural translation.
Read the Original
This page is a summary of: Statistical versus neural machine translation – a case study for a medium size domain-specific bilingual corpus, Poznan Studies in Contemporary Linguistics, June 2019, De Gruyter,
DOI: 10.1515/psicl-2019-0018.
You can read the full text:
Contributors
The following have contributed to this page