A time-series-based model to detect homogeneous regions of residents' dynamic living habits

The identification of homogeneous regions representing the dynamic living habits of residents has long been a central focus in human activity research. Although extensive studies have been conducted in this area, the majority of these efforts have concentrated on static residential behaviors, neglecting the temporal fluctuations in urban residents' activities throughout the day. This study introduces a novel model that integrates facility operation data with a time-series-based regionalization approach to detect residents' dynamic living habits at the block scale and identify homogeneous regions. The model first transforms textual data into temporal Origin-Destination (OD) flows and introduces the concept of an "urban schedule" to represent the Temporal OD of various resident activities within a specific unit. Building on this foundation, the urban schedule is subsequently converted into operating probability curves over time, revealing the spatial structure of residents' dynamic living habits. Our findings demonstrate that the homogeneous regions of residents' dynamic living habits not only reflect the intensity of human activity but also show a significant correlation with distinct urban functional areas, highlighting their importance for urban planning and development.