The paper introduces a novel method for extracting community structures in large networks, focusing on modularity optimization. The method is designed to be highly efficient in terms of computation time while maintaining good quality in community detection, as measured by modularity. The authors demonstrate the effectiveness of their algorithm through applications to a Belgian mobile phone network with 2.6 million customers and a web graph with 118 million nodes and over one billion links. The algorithm is divided into two phases: the first phase optimizes modularity by allowing local changes in community assignments, and the second phase aggregates these communities to form a new network of communities. This process is repeated iteratively until no further improvement in modularity is possible. The method is shown to outperform other community detection algorithms in terms of both speed and modularity, and it successfully identifies hierarchical community structures. The paper also discusses the limitations and potential improvements of the algorithm, emphasizing its ability to handle networks of unprecedented size and its potential for uncovering complex network structures.The paper introduces a novel method for extracting community structures in large networks, focusing on modularity optimization. The method is designed to be highly efficient in terms of computation time while maintaining good quality in community detection, as measured by modularity. The authors demonstrate the effectiveness of their algorithm through applications to a Belgian mobile phone network with 2.6 million customers and a web graph with 118 million nodes and over one billion links. The algorithm is divided into two phases: the first phase optimizes modularity by allowing local changes in community assignments, and the second phase aggregates these communities to form a new network of communities. This process is repeated iteratively until no further improvement in modularity is possible. The method is shown to outperform other community detection algorithms in terms of both speed and modularity, and it successfully identifies hierarchical community structures. The paper also discusses the limitations and potential improvements of the algorithm, emphasizing its ability to handle networks of unprecedented size and its potential for uncovering complex network structures.