Natural language processing for social inclusion: a text simplification architecture for different literacy levels

Caroline GasperinErick MazieroLucia SpeciaThiago PardoSandra Maria Aluísio

Text simplification is a research area of Natural Language Processing, whose goal is to maximize text comprehension through simplification of its linguistic structure. This paper presents our approach for Brazilian Portuguese text simplification. As people have different literacy levels, we take that into account when generating simplified texts. We propose an architecture for text simplification composed by two layers: the first is a machine-learning system who learns from manually simplified texts the appropriate degree of simplification according to a given literacy level; and the second is a rule-based system that executes the actual simplification of the sentences, following the recommendations from the first layer.

Caso o link acima esteja inválido, faça uma busca pelo texto completo na Web: Buscar na Web

Biblioteca Digital Brasileira de Computação - Contato:
     Mantida por: