The paper by Leicht and Newman addresses the problem of identifying community structures in directed networks, a common approach in previous literature being to ignore edge directions and apply methods designed for undirected networks. However, this approach discards valuable information contained in edge directions. The authors propose a method that generalizes the widely used modularity function to incorporate edge directions, allowing for the discovery of communities by maximizing modularity over possible network divisions. This is achieved using an algorithm based on the eigenvectors of the modularity matrix. The method is shown to outperform previous methods on various test networks, both real and computer-generated. The paper includes examples demonstrating the effectiveness of the directed modularity algorithm in uncovering meaningful structures in networks where undirected methods fail, such as in a word network and a random network with specific edge directions. The authors recommend using the directed algorithm for analyzing both edge placement and direction in directed graphs.The paper by Leicht and Newman addresses the problem of identifying community structures in directed networks, a common approach in previous literature being to ignore edge directions and apply methods designed for undirected networks. However, this approach discards valuable information contained in edge directions. The authors propose a method that generalizes the widely used modularity function to incorporate edge directions, allowing for the discovery of communities by maximizing modularity over possible network divisions. This is achieved using an algorithm based on the eigenvectors of the modularity matrix. The method is shown to outperform previous methods on various test networks, both real and computer-generated. The paper includes examples demonstrating the effectiveness of the directed modularity algorithm in uncovering meaningful structures in networks where undirected methods fail, such as in a word network and a random network with specific edge directions. The authors recommend using the directed algorithm for analyzing both edge placement and direction in directed graphs.