Understanding Residents’ Behaviors in a Housing Estate by Passive WiFi Sensing and Data Mining

摘要

Understanding residents' behaviors in a housing estate helps with the better management of the estate and the better design of future estates. Passive WiFi sensing, by collecting WiFi probe requests sent from mobile devices, offers a better way to conduct such studies compared with other methods due to little interference, larger coverage, lower cost, and more information on people’s movement. However, there are not many existing studies focused on studying housing estate leveraging the power of passive WiFi sensing. In this work, we collect data on residents' appearance and movement inside a housing estate through passive WiFi sensors. Afterward, data mining techniques, including visualization, linear regression analysis, and hierarchical agglomerative clustering, are exploited to extract insights on residents' behaviors. The studied aspects of residents' behaviors include daily activeness patterns, factors that affect the residents' activeness, and movement patterns between different parts of the estate.