What drives consumptive mobility patterns? Evidence from mobile internet usage data

Consumptive mobility, such as shopping and entertainment activities, which are discretionary and flexible travel behavior, are influenced by the socioeconomic and spatial attributes of heterogeneous individual travelers. But limited by the difficulty of collecting large-scale granular individual traveler profile attribute datasets, achieving a comprehensive understanding of the intricate determinants consumptive mobility patterns remains challenging, even when relying on a complete sample. In this study, we organize our analysis by representative shopping mall, and leverage precious mobile internet usage data, that containing the preferred websites and applications (apps) visited by mobile phone carrier. And then, they are integrated into portraits of traffic analysis zones (TAZs) that travel to shopping malls. To examine the determinants of travel volumes to shopping malls, we pre-select multiple key elements and employ a partial least squares (PLS) regression model. In our empirical studies, we test our framework in the real Suzhou, China, using a mobile phone dataset that provides comprehensive spatiotemporal coverage. This dataset includes the positioning data and socioeconomic data estimated based on internet usage preferences of the same users. Our findings reveal that distance, entertainment, dining, and shopping have a significantly higher impact on travel volumes compared to social activities, learning, travel, age, and gender.