Table 3.

Results of ML and DL models

Model name	tp	fp	tn	fn	f1	Acc (%)
Decision Trees	11,229	8690	2540	41	0.72	61.2
XGB	8895	4007	7223	2375	0.74	71.6
KNN	8867	6411	4819	2403	0.67	60.8
RNN	8544	4050	7180	2726	0.71	69.9
LSTM	4540	984	10,246	6730	0.65	69.5
Bi-LSTM	5223	1299	9931	6047	0.66	70.3
BERT	9252	1515	9715	2018	0.84	84.3
BERTweet	9564	1053	10,177	1706	0.88	87.7%

tp true positives, fp false positives, tn true negatives, fn false negatives, f1 f1-score, acc accuracy)