Silvia Maria Wanderley Moraes,  Vera Lúcia Strube de Lima. 

This paper presents an investigation about hierarchical text categorization using the k-Nearest Neighbor algorithm to classify a large corpus in Portuguese, the PLN-BR CATEG corpus. Besides describing the main difficulties to carry on the experiments and to analyze the obtainedresults, this work empirically studies the influence of some of the parameters in the categorization process.

http://www.lbd.dcc.ufmg.br/colecoes/til/2007/0012.pdf

Processando consulta.