Paper
Experiments on language normalization for spanish to english machine translation
Abstract
We describe a trained system we have built, using freely available statistical packages and present experiments specifically designed to improve the performance of the system in the Spanish into English task. We performed both qualitative and quantitative evaluations that show better perceived translation quality as well as better BLEU and NIST scores. © AEPIA.
Related
Conference paper
Spottune: Transfer learning through adaptive fine-tuning
Conference paper
Incorporating behavioral constraints in online AI systems
Conference paper