Revista: | Computación y sistemas |
Base de datos: | |
Número de sistema: | 000560526 |
ISSN: | 1405-5546 |
Autores: | Silva, Samuel Caetano da1 Ferreira, Thiago Castro2 Ramos, Ricelli Moreira Silva1 Paraboni, Ivandré1 |
Instituciones: | 1Universidade de São Paulo, Escola de Artes, Ciencias e Humanidades, Sao Paulo. Brasil 2Universidade Federal de Minas Gerais, Escola de Belas Artes, Minas Gerais. Brasil |
Año: | 2020 |
Periodo: | Jul-Sep |
Volumen: | 24 |
Número: | 3 |
Paginación: | 1179-1188 |
País: | México |
Idioma: | Inglés |
Tipo de documento: | Artículo |
Resumen en inglés | Computational models of hate speech detection and related tasks (e.g., detecting misogyny, racism, xenophobia, homophobia etc.) have emerged as major Natural Language Processing (NLP) research topics in recent years. In the present work, we investigate a range of alternative implementations of three of these tasks - namely, hate speech, aggressive behavior and target group recognition -by presenting a number of experiments involving different learning methods, including regularized logistic regression, convolutional neural networks (CNN) and deep bidirectional transformers (BERT), and using word embeddings, word n-grams, character n-grams and psycholinguistics-motivated (LIWC) features alike. Results suggest that a purely data-driven BERT model, and to some extent also a hybrid psycholinguisticly informed CNN model, generally outperform the alternatives under consideration for all tasks in both English and Spanish languages. |
Disciplinas: | Ciencias de la computación |
Palabras clave: | Inteligencia artificial |
Keyword: | Natural language processing, Hate speech, Aggressive language detection, Artificial intelligence |
Texto completo: | Texto completo (Ver HTML) Texto completo (Ver PDF) |