Desafíos para la traducción automática en lengua española: combinación de métodos y ontologías. Un análisis del servidor de traducción del Instituto Cervantes

The creation and improvement of automatic translation tools and technologies has been irregular. The application and purpose of some of these tools is still problematic, in particular, while implemented without linguistic resources like dictionaries or parallel corpora. In this paper, I explore some fundamental dimensions within texts that should be taken into account for the design and implementation of any automatic translator. I critically consider the theoretical approaches and methods applied for the design of automatic translation software by analysing different texts translated into spanish by the Instituto Cervantes online translation service. The purpose of this succinct study is to illustrate the challenges put by translation and suggest some answers related to the combination of methods and tools; the design and implementation of ontologies and formal rules based not only on linguistic information but also on knowledge; and the effective interaction of the machine and the human translator.