Comparison of attributed network clustering approaches to analysing Seoul public bike stations

Abstract

Network clustering is a fundamental task that discovers innate communities or groups in networks. Hence, network clustering methods such as spectral clustering and regularized spectral clustering have been applied in a wide range of realms. On top of a network structure, it is known in social network analysis that incorporates information from each vertex can be beneficial. This has led to the development of a series of attributed network clustering algorithms that utilize not only network connectivity but also vertex covariates in order to uncover latent clusters. This paper compares the performance of state‐of‐the‐art attributed network clustering approaches focused on detecting clusters of Seoul public bike stations. The data set consists of trip information over the bike station network in 2019. Spatial information about the bike stations is posed as vertex attributes. We show that certain attributed network clustering methods are well suited to detecting explainable clusters of bike rental stations. The results can help bike‐sharing operators better understand system usage and learn how to improve service quality in the existing system.