Informação e Critério de Relevância:podem as máquinas aprender a diferenciar contextos?

• Autor(a) principal: Rodrigues, Mariana Vitti [UNESP]
• Data de Publicação: 2011
• Outros Autores: Gonzalez, Maria Eunice Quilici [UNESP]
• Tipo de documento: Artigo
• Idioma: por
• Título da fonte: Repositório Institucional da UNESP
• Texto Completo: http://www.marilia.unesp.br/Home/RevistasEletronicas/FILOGENESE/MarianaVittiRodrigues.pdf; http://hdl.handle.net/11449/115237

The aim of this paper is to investigate the notion of criteria of relevance with the help of the concept of information characterized by Charles Sanders Peirce. For criteria of relevance we mean a process of choosing the best action in a context. To this end, we will introduce questions relating to the study of mind, especially those concerning the debate between mechanism X non-mechanism. In an attempt to shed light on issues concerning the nature of mind, we introduce the characterization, proposed by Peirce, the concept of information. Differentiating the concept of genuine information and of degenerate information, arguing that the detection of genuine information allows to characterize the relevant actions in a given context. Initially, we will organize our work questioning whether a machine can have autonomous criteria of relevance, that is, if it can learn different criteria formulated by your programmer. Investigating whether an artificial being can learn to differentiate between contexts and therefore act in the best possible way on each occasion. Finally, we will make an balance of Peirce's contributions to this theme.