查看文章

Identifying the urban community structure is important for promoting smart city. In contrast to community detection in social network, the urban community construction involves both the information of population flow a nd t he distribution of public services. However, few methods take into account the two factors simultaneously in a reasonable manner. This causes several problems such as isolated nodes and geographical fragmentation in urban planning. To address the challenges of concerning the population flow a nd p ublic s ervices distribution in an unified framework, we formulate the identification problem of urban community structure as graph modularity optimization with constraints on the public services distribution. The combination of hierarchical clustering and node level adjustment strategies is applied, to capture the required information of community structure. In our framework, The attributed graph …