Fig. 4

Example partition of the Enron data set via approximate spectral maximization of Eq. (9). (Left): The adjacency matrix, arranged in order of the partition. Colors give the value of logW_ij+1; log-scaling was chosen for readability. (Right): Visualization of the network. Directed edges have been made undirected. The darkness of each edge corresponds to its weight. The color of each node corresponds to listed job title. In this experiment, k=4, and the kernel used is the information flow kernel R of Eq. (??). The modularity of the shown partition is 0.506