Learning analytics-scientific description and heuristic validation of languages NLG

Educators do not have to be mere "Translators" as manufacturers of algorithms for teaching in the infosphere, but must modulate it for everyone's purposes. This work aims to validate the moments of transformation, through which, over the centuries, the mathematical sciences, with the help of philosophy, have promoted the Natural Language Generation to formal models. The starting hypothesis intends to corroborate an epistemological framework, which assigns logical-mathematical reasoning to mental processes following four models: Chomsky, who, with descriptive grammar marks a new model for rewriting languages; Gross, who, with the relationship between linguistics, computer science, and mathematics generates a relationship concerning a strongly transdisciplinary domain, in which linguistics realizes computer-based models and procedures; Silberzstein's Nooj system for processing, describing and analyzing INLG fixed sentences. The focal part of the research is the work of comparison that the team carried out to validate the development of languages according to the Transformational Analysis of Direct Transitive by M. Silberzstein and the lexicon-grammar; probabilistic calculus, according to the Probabilistic latent semantic Analysis.