Feature selection for multi-label text data: An ensemble approach using geometric mean aggregation | Publicación