| Resumen |
Web published news written in the Spanish language, were analyzed by using categories that are related to its content, such as: ’Culture’, ’Sports’ and ’Finances’, or they are classified very generally as is the case of ’National’ or ’International’. The large content of documents generated the need to provide the user with an analysis of such documents, particularly in circumstances where in search engines are involved. First of all, a pre-process was applied to allow the mining of texts, which includes the lemmatization, homologation of synonyms and representation of documents with a Boolean method. This pre-process also includes a dimensional reduction of the obtained matrix. Secondly, different classification methods were applied to compare their performance in order to find the one that best assigns the category to the news. |
