Training and Fine-Tuning NMT Models for Low-Resource Languages Using Apertium-Based Synthetic Corpora | Publicación