Elhuyar technology wins internationally

2020/09/29 Elhuyar Zientzia

• Sariak
• Adimen artifiziala
• COVID-19
• Elhuyarren berriak

400

Specifically, the field of videomedicine is due to the special effort made by Elhuyar in the face of the pandemic caused by the SARS-CoV-2 virus (causing the disease COVID-19), to which research and development has led. In this sense, he has developed an automatic neuronal translator for health (English-Basque), which has been ranked first in the international competition Biomedical Translation Task.

In addition, in a contest organized by the US NIST Institute, he has presented a neuronal search engine for covid-19 related queries in scientific articles and has ranked second among the best systems of 28 competitors.

At this time, Elhuyar is working on a grammar checker based on neural networks for texts written in Basque. Well, on September 23, at the congress of the Spanish Society for the Processing of Natural Language (SEPLN) presented the details of our research, obtaining the award for best article.

The three winning awards show that Elhuyar's research and the resulting technology are leading and demonstrate that these technologies can be successfully applied in fields such as biomedicine and language use.

Elhuyar is not new in the field of artificial intelligence technologies, has a long history in research and development. He has already published more than 100 scientific articles, developed more than 10 solutions and socialized customized projects.

The main lines are: machine translation (itzultzailea.eus), speech processing (voice synthesis, machine transcription), search systems, knowledge extraction systems and language tools (e.g. automatic corrector).

In all these areas of research it uses deep neural networks, obtaining technologies of maximum efficiency. The technologies developed by Elhuyar are applicable to different sectors: marketing, journalism, banks, digital humanities... In addition, it proposes solutions not only for Basque, but also for other languages.