Emotion Recognition and Classification of Film Reviews Based on Deep Learning and Multimodal Fusion

Table 1

Main parameter settings of the SA-BLCNN model.

Parameter name	Value
Word vector dimension	100
Sliding window size	3, 4, 5
Activation function	ReLU
Pooling method	Max-pooling
Dropout	0.5
Epoch	50
Learning rate	0.01
Loss function	Crossentropy
Optimization function	Adam
Hidden layer size	128
Batch size	32