Table 3. Experimental results for Twitter dataset.

Twitter	Machine learning					Deep learning			Proposed technique
Dataset	NB	LR	SVC	DT	RF	LSTM	Bi-LSTM	GRU	Focal loss
Accuracy	85.3	90.7	92	90.3	90.6	90.7	91.8	91.3	99.00
Precision	85.1	80.3	81.5	69.8	83.6	67,5	72.06	73.3	86.72
Recall	5.6	52.5	61.9	65.6	48.4	71.8	72.8	65.2	74.67
F1 Score	10.4	63.5	70.4	67.6	61.3	69.9	72.4	69.0	73.88
Training time (s)	0.02	1.2	44.6	7.5	13.5	526.43	906.32	612.24	516.784
Testing time (s)	0.11	0.16	5.3	0.11	0.4	4.67	8.64	5.27	4.814

Notes.

The best performing results are shown in bold.