Graph neural network is a new intelligent algorithm in recent years. It combines deep learning algorithm and graph computing algorithm to learn from each other to achieve better cognition and problem processing. It is widely used in important fields such as search, recommendation, risk control and so on. Graph neural network (GNN) is becoming more and more popular in various fields, including social networks, knowledge maps, recommendation systems, and even life sciences. The powerful function of GNN in modeling the dependency relationship between nodes in the graph has made a breakthrough in the research field related to graph analysis. Neural network workflow: according to the preliminary understanding, neurons are connected end to end in the figure and realize data transmission. The output of the previous neuron will become the input of the next neuron.