Discovering the influence of facility distribution on lifestyle patterns in urban populations

Abstract

The spatial structures of cities defined by population distribution, distribution of facilities, and mobility have a significant impact on lifestyles of residents and their wellbeing. In this study, we analyze millions of mobile phone data points to infer significant sequences of visited facilities by individuals, cluster people with similar patterns of life activity sequences, and define lifestyles based on the patterns in each cluster. We find that lifestyles of a large number of people can be captured using a small set of activity sequences, while a small portion of populations have lifestyles with a variety of activity sequences. Facility proximity in spatial constraints is positively correlated with the volume of human movements, and is a significant factor in formation of the majority of lifestyle patterns. Differences in facility proportions between two neighborhoods contribute to cross-neighborhood travels for life activities, but its effect could be mediated by the geographical distances between neighborhoods. Our findings demonstrate that the widely studied scaling laws in these areas are not independent but rather connected through a deeper underlying reality, which has important implications for urban planning and city management policies to enhance equal accessibility.

Keywords

• Human mobility
• Population lifestyles
• Facility distribution
• Urban computing