Neural networks at the human mind: translator's hidden layers

摘要

This article is based on training process for students to become specialized translators and post-editors. Any language mediator working within specialized communication may use a graphic representation to outline the contents of a text in a particular domain. The aim of this graphic depiction is to draw mental structures that distinguish the main concepts from the subordinates, while linking them to their linguistic referents of the different languages. This conceptual diagram can be seen as a neural network: there is an input layer (terms in a text), a hidden layer where the student comes into the upper ordinate and subordinate layers and hierarchies (mind), and an output layer (the diagram bridging equivalences). Young students at Translation Faculty must front the fact that it is not only language skills what they need, but an objective conceptual diagram on domains they have no previous idea of the subject.