Table 3. The summary of dataset description.

Data source	Twitter
Data collection approach	Automatic retrieval of tweets using keywords (ARTK)
Language of tweets	English
Data classes	Sarcastic and non-sarcastic
Search period	Between June 2019 and September 2019
Sarcastic data volume	15,000
Non-sarcastic data volume	14,913
Total volume of data	29,931
Annotation	Self-annotated by tweet owner
Sarcastic annotation	‘#sarcasm’ or ‘#sarcastic’
Non-sarcastic annotation	#notsarcasm or #notsarcastic or without any hashtag