Data-driven behavioral analysis and applications: A case study in Changchun, China

The mobile phone data have become crucial in behavioral analysis to detect habits of human mobility and reveal rules of behaviors. Previously, questionnaires were often used to identify urban functional areas, with vast labor and poor timeliness. To address the issue, this paper applies data-driven behavioral analysis to identify functional areas for governments to construct urban design, offer site selection and manage transportation. Moreover, data-driven behavioral analysis can also be applied in student behaviors to help schools adjust facility arrangements, develop learning efficiency and provide high-quality services. Therefore, based on mobile phone data in Changchun, this paper utilizes a two-stage clustering method combining human mobility to identify urban functional areas, including business, working, residential and low passenger-flow areas. The interesting finding is that local prosperity in Changchun is prominent and the proportion of low passenger-flow areas can reflect the development level. Furthermore, this paper compares student behaviors in three schools, which shows each school varies in distribution features of students. Experiments provide enlightening insights to reveal the spatial structure of cities and comprehend the living state of students.(c) 2022 Elsevier B.V. All rights reserved.