“…Previous work mainly focuses on using machine learning techniques to build text classifiers. Several methods have been proposed in the literature for the construction of document classifiers, such as decision trees [5], Support Vector Machines [13], Bayesian classifiers [24], hierarchical text classifiers [19], [11], [9], [20], [26], [12], [21], [7], [17]. The main commonality in previous methods is that their classification accuracy depends on a training phase, during which statistical techniques are used to learn a model based on a labeled set of training exampled.…”