Understanding human behavior from aggregated smartphone signals and land use information

Authors: Ta-Chien Chan*, Center for Geographic Information Science, Research Center for Humanities and Social Sciences, Academia Sinica, Taipei, Taiwan, Tzu-Yu Lin, Center for Geographic Information Science, Research Center for Humanities and Social Sciences, Academia Sinica, Taipei, Taiwan
Topics: Human-Environment Geography, Geographic Information Science and Systems, Spatial Analysis & Modeling
Keywords: Human dynamics, demographic behavior, data visualization, big data
Session Type: Poster
Day: 4/5/2019
Start / End Time: 8:00 AM / 9:40 AM
Room: Lincoln 2, Marriott, Exhibition Level
Presentation File: No File Uploaded


An understanding of the real spatial distribution and dynamic flow of human beings is important for both the social sciences and city governance. If we can use massive aggregated mobile phone data, it not only overcomes the sampling coverage issue but also makes it feasible to explore the spatial distribution of different demographic behaviors and their relationship to land use. This study collaborated with one telecommunication company in Taiwan to obtain data from mobile internet access users via 38,317 service grids in Taipei City and New Taipei City. The study period was two weeks with spatial resolution of 250 meters and temporal resolution of 10 minutes. In addition, we also utilized national land use survey data. Then, we used hierarchical clustering to model the relationship between human distribution and land use by Python. The results found that we could primarily classify human movement patterns into two kinds, including one pattern in the morning and night and the other one at noon. In the former, we found that most human activities clustered in residential and commercial land use. In the latter, we found most human activities clustered at land used for service, production services and recreational sites. Elderly people seldom appeared at locations used for service and production services, and more frequently appeared at public land locations such as parks and hospitals. Through data mining and classification of that information, we can identify different demographic behaviors and use them as basic information for social sciences research and city planning.

Abstract Information

This abstract is already part of a session. View the session here.

To access contact information login