LX-AuthorialStyle's Documentation
LX-AuthorialStyle
LX-AuthorialStyle is an online service for the recognition of the authorial style under which a given Portuguese text is written, among 80 possible authorial styles, corresponding to an equal number of authors whose works are included in the BDCamões Corpus of literary texts (XVI to XXI centuries). This service can also be used to find the author of an excerpt of a work provided this work is in the corpus.
This service is based on automatic text processing tools. Hence, the results it returns may not be always totally correct. Its high accuracy rate though allows it to provide a useful service given its aim.
This service was developed and is maintained at the University of Lisbon University of Lisbon by NLX-Natural Language and Speech Group of the Department of Informatics.
Features and evaluation
The classifier is based on GPorTuguese-2, a "GPT-2 small" deep neural network with 124.4 million parameters. Its accuracy is 93%.
Authors and works
These are the authors included in BDCamões Corpus:
| author | works | words |
|---|---|---|
| Agustina Bessa-Luís | 7 | 378,522 |
| Alexandre Herculano | 8 | 173,851 |
| Alfredo Margarido | 1 | 9,646 |
| Almeida Garrett | 4 | 123,208 |
| Amadeu Lopes Sabino | 1 | 4,621 |
| Antero de Quental | 3 | 54,211 |
| António Botto | 1 | 2,770 |
| António Feliciano de Castilho | 1 | 5,385 |
| António José da Silva | 1 | 23,877 |
| Aquilino Ribeiro | 6 | 46,295 |
| Armando Silva Carvalho | 1 | 2,131 |
| Augusto Abelaira | 1 | 3,129 |
| Bernardo Gomes Brito | 1 | 8,871 |
| Bernardo Santareno | 1 | 8,247 |
| Brito Camacho | 1 | 4,980 |
| Camilo Castelo Branco | 7 | 177,012 |
| Conde de Ficalho | 2 | 5,521 |
| D. Francisco Manuel de Melo | 1 | 18,591 |
| David Mourão-Ferreira | 1 | 5,623 |
| Eça de Queirós | 10 | 273,011 |
| Fernando Cabral Martins | 2 | 1,798 |
| Fernando Pessoa | 1 | 5,154 |
| Fernando Venâncio | 1 | 2,855 |
| Fernão Lopes | 1 | 36,410 |
| Fernão Mendes Pinto | 2 | 19,004 |
| Ferreira de Castro | 1 | 4,347 |
| Fialho D'Almeida | 5 | 92,185 |
| Francisco Maria Bordalo | 1 | 13,395 |
| Gil Vicente | 6 | 21,068 |
| Gonçalo M. Tavares | 3 | 1,773 |
| Hélia Correia | 1 | 2,567 |
| Jacinto Lucas Pires | 1 | 2,895 |
| Jaime Rocha | 1 | 3,801 |
| Jerónimo Osório de Castro | 1 | 8,319 |
| João Braz de Oliveira | 1 | 5,318 |
| João Vaz | 1 | 8,964 |
| Joaquim Canas Cardim | 1 | 4,443 |
| Joaquim Paço D'Arcos | 1 | 12,521 |
| Joaquim Pedro Celestino Soares | 1 | 10,218 |
| Jorge de Sena | 5 | 37,684 |
| José Cardoso Pires | 1 | 6,447 |
| José de Almada Negreiros | 3 | 14,326 |
| José Luandino Vieira | 2 | 21,089 |
| José Martins Garcia | 1 | 6,946 |
| José Régio | 1 | 10,836 |
| José Rodrigues Miguéis | 2 | 17,934 |
| Júlio Dantas | 2 | 6,774 |
| Júlio Dinis | 5 | 528,249 |
| Lídia Jorge | 2 | 13,942 |
| Luís de Camões | 1 | 146,821 |
| Luísa Costa Gomes | 3 | 16,248 |
| Luísa Dacosta | 1 | 9,798 |
| Manuel de Arriaga | 1 | 21,686 |
| Manuel Maria Barbosa du Bocage | 7 | 19,622 |
| Manuel Teixeira Gomes | 5 | 26,160 |
| Maria Gabriela Llansol | 1 | 2,373 |
| Maria Leonor Buescu | 1 | 32,097 |
| Maria Ondina Braga | 1 | 4,927 |
| Maria Teresa Horta | 1 | 1,498 |
| Maria Velho da Costa | 1 | 1,020 |
| Mário Cláudio | 1 | 578 |
| Mário de Carvalho | 5 | 22,235 |
| Mário de Sá-Carneiro | 1 | 2,218 |
| Mário Henrique Leiria | 1 | 731 |
| Maximiano Lemos Júnior | 1 | 6,263 |
| Nun'Álvares de Mendonça | 1 | 17,568 |
| Nuno Júdice | 2 | 3,850 |
| Oliveira Martins | 3 | 334,693 |
| Padre António Vieira | 1 | 12,038 |
| Pêro Vaz de Caminha | 1 | 9,395 |
| Ramalho Ortigão | 6 | 239,252 |
| Raul Brandão | 3 | 69,207 |
| Ruben A. | 1 | 5,878 |
| Rui de Pina | 8 | 219,031 |
| Sophia de Mello Breyner | 1 | 6,711 |
| Teófilo Braga | 5 | 227,856 |
| Teresa Veiga | 1 | 8,056 |
| Tomaz de Figueiredo | 1 | 4,308 |
| Tomaz Vieira da Cruz | 1 | 4,224 |
| Trindade Coelho | 18 | 127,166 |
| Venceslau de Moraes | 2 | 43,776 |
| Vergílio Ferreira | 2 | 6,247 |
| Vitorino Nemésio | 4 | 41,648 |
The list of the respective works can be found here. These works can be read at Biblioteca Digital Camões.
Authorship
LX-AuthorialStyle was developed by João Silva and Rodrigo Santos under the coordination of António Branco, at NLX-Natural Language and Speech Group.Contact us
Contact us using the following email address: 'nlx' concatenated with 'at' concatenated with 'di.fc.ul.pt'.
Why LX-AuthorialStyle?
LX because LX is the shorthand form Lisboners often use to refer to their hometown.
License
No fee, attribution, all rights reserved, no redistribution, non commercial, no warranty, no liability, no endorsement, temporary, non exclusive, share alike.
The complete text of this license is here.