Linguistic literature on irony discusses sarcasm as a form of irony characterized by its biting nature and the intention to mock a victim. This particular trait makes sarcasm apt to convey hate speech and not only humour. Previous works on abusive language stressed the need to address ironic language to lead the system to recognize correctly hate speech, especially in spontaneous texts, like tweets [13]. In this context, our main hypothesis is that information about the presence of sarcasm could help to improve the detection of hateful messages, especially when they are camouflaged as sarcastic. To corroborate this hypothesis: i) we perform analysis on HaSpeeDe20_ext, an Italian corpus of tweets about the integration of cultural minorities in Italy, ii) we carry out computational experiments injecting the knowledge of sarcasm in a system of hate speech detection, and iii) we adopt strategies of validation in terms of performance and significance of the obtained results. Results confirm our hypothesis and overcome the state of the art.

When Sarcasm Hurts: Irony-Aware Models for Abusive Language Detection

Frenda S.;Patti V.;Rosso P.
2023-01-01

Abstract

Linguistic literature on irony discusses sarcasm as a form of irony characterized by its biting nature and the intention to mock a victim. This particular trait makes sarcasm apt to convey hate speech and not only humour. Previous works on abusive language stressed the need to address ironic language to lead the system to recognize correctly hate speech, especially in spontaneous texts, like tweets [13]. In this context, our main hypothesis is that information about the presence of sarcasm could help to improve the detection of hateful messages, especially when they are camouflaged as sarcastic. To corroborate this hypothesis: i) we perform analysis on HaSpeeDe20_ext, an Italian corpus of tweets about the integration of cultural minorities in Italy, ii) we carry out computational experiments injecting the knowledge of sarcasm in a system of hate speech detection, and iii) we adopt strategies of validation in terms of performance and significance of the obtained results. Results confirm our hypothesis and overcome the state of the art.
2023
Proceedings of the 14th International Conference of the Cross-Language Evaluation Forum for European Languages, CLEF 2023
grc
2023
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Springer Science and Business Media Deutschland GmbH
14163
34
47
978-3-031-42447-2
978-3-031-42448-9
Abusive Language Detection; Irony Detection; Multi-Task Learning
Frenda S.; Patti V.; Rosso P.
File in questo prodotto:
File Dimensione Formato  
C18_2023___.pdf

Accesso riservato

Tipo di file: PDF EDITORIALE
Dimensione 10.07 MB
Formato Adobe PDF
10.07 MB Adobe PDF   Visualizza/Apri   Richiedi una copia

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/2318/1945387
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 0
  • ???jsp.display-item.citation.isi??? ND
social impact