Table 4 Classification report of different models for sentiment analysis.

Class label	Measures	Logistic regression	CNN	Bi-LSTM	RoBERTa	BERT	Adapter-BERT
Positive	Precision	0.42	0.62	0.05	0.22	0.23	0.38
	Recall	0.37	0.80	0.23	0.28	0.20	0.25
	F1-Score	0.39	0.69	0.09	0.24	0.21	0.31
	Support	614	2379	109	469	469	469
Negative	Precision	0.66	0.17	0.24	0.37	0.41	0.40
	Recall	0.76	0.08	0.46	0.37	0.36	0.39
	F1-Score	0.71	0.11	0.32	0.37	0.39	0.43
	Support	2058	542	282	542	542	542
Mixed_Feelings	Precision	0.27	0.28	0.92	0.74	0.73	0.82
	Recall	0.21	0.18	0.66	0.69	0.77	0.78
	F1-Score	0.24	0.22	0.77	0.71	0.75	0.73
	Support	601	686	3287	2379	2379	2379
Unknown_State	Precision	0.47	0.15	0.29	0.40	0.46	0.46
	Recall	0.40	0.09	0.51	0.42	0.45	0.39
	F1-Score	0.43	0.11	0.37	0.41	0.46	0.44
	Support	803	469	398	686	686	686

Quick links

Search