Tanz-Indicator: A Novel Framework for Detection of Perso-Arabic-Scripted Urdu Sarcastic Opinions

Table 3

Performance comparison of lexicon model with machine learning models.

Performance matrices	Lexicon	Naïve Bayes	SVM
Precision	0.879	0.828	0.777
Recall	0.696	0.201	0.204
-measure	0.776	0.324	0.323
Accuracy	0.720	0.652	0.601