Evaluating machine translation in a low-resource language combination: Spanish-Galician.
María do Campo Bayón, Pilar Sánchez-Gijón · Dipòsit Digital de Documents de la UAB (Universitat Autònoma de Barcelona) · 2019
This paper reports the results of a study designed to assess the perception of adequacy of three different types of machine translation systems within the context of a minoritized language combination (Spanish-Galician). To perform this evaluation, a mixed design with three different metrics (BLEU, survey and error analysis) is used to extract quantitative and qualitative data about two marketing letters from the energy industry translated with a rulebased system (RBMT), a phrase-based system (PBMT) and a neural system (NMT). Results show that in the case of low-resource languages rule-based and phrase-based machine translations systems still play an important role.