Automatic classification of new articles in Spanish

Abstract

We apply machine learning techniques to the automatic classification of news articles from the local newspaper La Capital of Rosario, Argentina. The corpus (LCC) is an archive of approximately 75,000 manually categorized articles in Spanish published in 1991. We benchmark on LCC three widely used supervised learning methods: k-Nearest Neighbors, Na¨ ve Bayes and Arti ficial Neural Networks, illustrating the corpus properties.