[Retracted] Network Public Opinion Risk Prediction and Judgment Based on Deep Learning: A Model of Text Sentiment Analysis

Table 4

Classification accuracy results after each model selects different activation functions.

Model	Positive data			Negative data
	Sigmoid	Relu	Tanh	Sigmoid	Relu	Tanh
BERT-CNN	0.7133	0.7053	0.7002	0.7205	0.7272	0.7232
BERT-Bi LSTM	0.6947	0.7025	0.6887	0.7212	0.7269	0.7215
BCBL	0.7176	0.7015	0.7041	0.7323	0.7234	0.7263