Network flow recognition method and device, and computer storage medium

The invention discloses a network flow recognition method and device, and a medium, and the method comprises the steps: preprocessing collected network data, extracting the feature information of eachnetwork session, and generating a flow log; based on the flow log, constructing a first-form flow graph in a preset time period; adding related edges for constructing the correlation between the server-side nodes in the first-form traffic graph, and generating a second-form traffic graph; training the graph neural network by utilizing the data of the second-form flow graph to generate a network flow recognition model; converting the unknown traffic into the second-form traffic graph and inputting the second-form traffic graph into a network traffic recognition model; enabling the network traffic recognition model to match the second form traffic graph of the unknown traffic with the graph mode learned by the network traffic recognition model; and when the matching degree is greater than apreset threshold, judging