Table 1. Experimental parameters and their set values.

Parameter	Set value
Regularization parameter ( $\lambda$)	0.01
Learning rate	0.1
Number of iterations (T)	100
Network depth (L, graph convolutional network parameter)	2
Node feature dimension (graph convolutional network parameter)	64
Range of initial values for embedding matrix	[−0.01, 0.01]
Initial value of user embedding matrix $\mathbf{P}$	Random initialization
Initial value of item embedding matrix $\mathbf{Q}$	Random initialization
Initial value of node embedding Z	Random initialization
Activation function (ReLU)	Yes
Loss function	Mean squared error
Optimizer	Adam
Batch size	128
Number of pre-training epochs	10
Total number of epochs	50
Early stopping rounds	5
Weight decay	0.0001
Number of negative samples	5
Embedding layer dimension	50
Dropout rate	0.5
Beta parameters for ADAM optimizer	[0.9, 0.999]