A few years ago machine translation began to take steps. Formerly predominated systems based on language rules (RBMT technology). Later they highlighted the systems based on statistics (SMT technology). In the last two years, systems based on deep neural networks are achieving qualitative improvements (NMT technology). In fact, today, machine translation research focuses primarily on deep learning.

Also in Euskal Herria. Within the Modela project, 5 entities collaborate in the research of translation systems based on neuronal networks for pairs of Spanish-Basque and Spanish-English languages: Ametzagaiña, UPV Ixa Group, Vicomtech/IK4, ISEA and Elhuyar.

To make this project known and learn more about the progress being made worldwide in neuronal machine translation, we organized an open day on 29 May. In addition to the members of the Modela project, we will have a special guest: Kyunghyun Cho, a renowned researcher at the University of New York. One of the most recognized experts in the field of neuronal machine translation and, in general, deep learnig, has received numerous awards, among which stand out the Google award granted last year.