Describir: Clustering by Similarity of Brazilian Legal Documents Using Natural Language Processing Approaches