Learning similarity metric to improve the performance of lazy multi-label ranking algorithms | Publicación