This article proposes a new method for the automatic recognition and correction of misspelled Spanish words called Ar2p-Text, inspired by the systematic theory of the human mind based on the recognition of patterns. Results are presented in several scenarios and a comparative evaluation with spell-checkers. The results show that Ar2p-Text is competitive, n the test scenarios. La Sección 3 hace una descripción formal de la arquitectura general de Ar2p-Text, su estructura de datos (módulos de reconocimiento de patrones) y su modelo