De agentes racionais a agentes semióticos: um estudo sobre a aplicação da semiótica na concepção de sistemas inteligentes.

Abstract

Rational agents are intelligent systems capable of building their own knowledge through the interaction with other agents. This systems are also capable of explaining their own decisions. In the present dissertation we study the evolution of such a system from the conception of a rational agent to a general experience learning system based on the Semiotics theory of Charles Sanders Peirce. For this purpose, firstly is shown the conception of the rational agent SAID, based on the Semi-Empirical Theory. By starting from SAID we develop a presentation of Semiotic in order to explore the structuring of experience based on signs. Secondly, a Semiotic Model of Reference for the conception of intelligent systems is shown. Finally, we develop a conceptual analysis of SAID under the Semiotics perspective, as well as na analysis of previous researches on intelligent systems based on the semiotic realm.