This study estimates living areas based on residents' actual travel patterns in Jeonju City, a small-to-medium- sized regional city, using mobile phone big data and community detection techniques. Existing living area plans have been established based on administrative boundaries, which often fail to reflect residents' actual mobility behaviors. To address this limitation, this study utilized floating population O/D data collected from LG U+ base stations, covering a total of 7,626,366 trips recorded in April 2024. The entire city was divided into 100×100m grid units.
Community detection was performed using the Leiden algorithm, with travel networks constructed for the total population as well as by age group, categorized into children and adolescents (under 19), youth (20~39), middle-aged (40~64), and elderly (65 and over). The modularity values ranged from 0.26 to 0.44 across all groups, confirming statistically meaningful community structures.
Results show that the travel-based living areas derived from community detection differ substantially from the five large living zones designated in Jeonju's Urban Master Plan. The total travel analysis identified six communities, with the northern and western living zones each subdivided into two communities, and one community spanning across the boundary between the southern and western zones. Age-specific analyses further revealed that adolescents formed eight communities while youth, middle-aged, and elderly groups each formed seven, indicating that spatial living area structures vary considerably by age group.
These findings suggest that living area plans based solely on administrative boundaries are insufficient to capture residents' actual mobility patterns. This study provides empirical evidence and foundational data for establishing travel pattern-based, age-responsive living area plans in regional cities.