BDB Bertsolaritzaren datu-basea

This work aims to study the narrative structure of Basque greeting verses from a text classification approach. We propose a set of thematic categories for the correct classification of verses, and then, use those categories to analyse the verses based on Machine Learning techniques. Classification methods such as Naive Bayes, k-NN, Support Vector Machines and Decision Tree Learner have been selected. Dimensionality reduction techniques have been applied in order to reduce the term space. The results shown by the experiments give an indication of the suitability of the proposed approach for the task at hands.