Table 5.

Hyperparameter values.

Hyperparameter	Value/Description
Text embedding	Dimension: 250
BLSTM layer	2 layers; 250 hidden units in each (Forward and backward)
Dense layer	3 layers; First 2 layers have 150 and 75 units respectively and the last one is output (Dense)
Drop-out rate	Word embedding: 0.3; Dense layer: 0.2 each;
Activation function	Conv1D, BLSTM, Dense: ReLU; Output dense layer: Sigmoid;
Adam optimizer	Learning rate 0.001–0.00001; $bet{a}_1=$ 0.8;
Validation	Training and validation split $=$ 80/20;
Epochs and batch	Epochs $=$ 50; batch size $=$ 68;