Finding the most influential nodes in complex networks is one of the open research issues. This problem can be divided into two sub-problems: (1) identifying the influential nodes and ranking them based on the individual influence of each node and (2) selecting a group of nodes to achieve maximum propagation in the network. In most of the previous articles, only one of these sub-issues has been considered. Therefore, this article presents a method to measure the spreading power of influential nodes in the network (the first sub-problem) and select the best group from them (the second sub-problem). In the proposed method, first, the input network is allocated to different communities. Then, the common neighbors and the degrees of the two end vertices of each edge are used to weigh the graph edges in each community. Next, in each of the communities, the nodes' propagation power is measured and ranked. Finally, a group of influential nodes is selected to start the propagation process.Eight data sets collected from real networks have been used for evaluation. The proposed method is compared with other previously known methods based on ranking accuracy, assigning different ranks to nodes, and calculating the amount of diffusion created in the network. The results show the proposed method's significant superiority over other methods in all test datasets.

Finding the most influential nodes in complex networks is one of the open research issues. This problem can be divided into two sub-problems: (1) identifying the influential nodes and ranking them based on the individual influence of each node and (2) selecting a group of nodes to achieve maximum propagation in the network. In most of the previous articles, only one of these sub-issues has been considered. Therefore, this article presents a method to measure the spreading power of influential nodes in the network (the first sub-problem) and select the best group from them (the second sub-problem). In the proposed method, first, the input network is allocated to different communities. Then, the common neighbors and the degrees of the two end vertices of each edge are used to weigh the graph edges in each community. Next, in each of the communities, the nodes' propagation power is measured and ranked. Finally, a group of influential nodes is selected to start the propagation process. Eight data sets collected from real networks have been used for evaluation. The proposed method is compared with other previously known methods based on ranking accuracy, assigning different ranks to nodes, and calculating the amount of diffusion created in the network. The results show the proposed method's significant superiority over other methods in all test datasets.