Mining the underlying relation network based on the communication actions

Information diffusion and virus propagation are fundamental processes, which achieve great attention in networks. While information spreads across the network, but the relations between the nodes, the underlying relation network, are usually unknown, and the authors only observe the time which transactions take place between nodes, called communication actions. Furthermore, the content of such transactions is unknown in many applications. They tackle these challenges by developing the NetMine algorithm to infer the relations between nodes and find out the underlying relation network, which is hidden from them. Given the times when nodes transmit a message to others (i.e. who sends a message to whom at some explicit time), they infer the optimal underlying relation network that best explains the observed communication actions. They demonstrate the effectiveness of their algorithm on synthetic and real networks, outperforming the art‐of‐date algorithms related to this problem.