This contribution presents an innovative tool called ‘sAIrcasm’ powered by Artificial Intelligence (AI). Challenges concerning the use of Artificial Intelligence and Natural Language Processing (NLP) in linguistics have been the subject of recent debate; one of the major challenges concerns the automatic detection of complex communicative phenomena, such as sarcasm. Several attempts have been made, but the complexity of the task has not yet led to the creation of an effective tool providing human-like results. This article presents a custom GPT model capable of identifying sarcasm and making pragmatic inferences based on context-dependent verbal and paraverbal cues. This study considers a Facebook video analyzed by both a custom artificial intelligence model and the free version of ChatGPT, providing insights into the broader implications of AI-based models in linguistic studies, with a specific focus on whether they can perform human-like cognitive processes.

Introducing sAIrcasm: A sample analysis of a custom Artificial Intelligence for linguistic and discursive sarcasm recognition

Alessia BATTISTA
2025-01-01

Abstract

This contribution presents an innovative tool called ‘sAIrcasm’ powered by Artificial Intelligence (AI). Challenges concerning the use of Artificial Intelligence and Natural Language Processing (NLP) in linguistics have been the subject of recent debate; one of the major challenges concerns the automatic detection of complex communicative phenomena, such as sarcasm. Several attempts have been made, but the complexity of the task has not yet led to the creation of an effective tool providing human-like results. This article presents a custom GPT model capable of identifying sarcasm and making pragmatic inferences based on context-dependent verbal and paraverbal cues. This study considers a Facebook video analyzed by both a custom artificial intelligence model and the free version of ChatGPT, providing insights into the broader implications of AI-based models in linguistic studies, with a specific focus on whether they can perform human-like cognitive processes.
File in questo prodotto:
Non ci sono file associati a questo prodotto.

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/11367/157640
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 1
  • ???jsp.display-item.citation.isi??? ND
social impact