Bruno Cunha Coutinho,  Jalmaratan Luís de Melo Macêdo,  Aroldo Rique Júnior,  Leonardo Vidal Batista. 

Many natural language processing techniques have been applied to the problem of authorship attribution, including neural networks, bayesian classifiers, and the Cusum method. E icent data compression algorithms, such as prediction by partial matching (PPM) and Lempel-Ziv algorithms have also been investigated. However, there seems to be very few researches in authorship attribution for Portuguese texts. This paper presents a method based on PPM-C compression algorithm for authorship attribution applied to brazilian literature texts.

http://www.lbd.dcc.ufmg.br/colecoes/til/2005/0020.pdf

Processando consulta.