Sandro J. Rigo,  José Palazzo M. de Oliveira,  Cristiano Barbieri. 

With the enormous quantity of documents that are now available on the Web, accessing and collecting desired data has become a difficult task that produce low quality results. The text classification permits a reduction of this problem. This work aims to describe an implementation of a text classification system with the use of linguistic information, described in a domain ontologywith the necessary information for the identification of both structure and concepts of documents associated to a specific class.

http://www.lbd.dcc.ufmg.br/colecoes/til/2007/0010.pdf

Processando consulta.