Unraveling the consumer geography from the review big data: A supply-demand duality perspective using store density and expenditure intensity

Consumer expenditure, a key indicator of socio-economic conditions and living standards, is a focus of multidisciplinary research. However, offline consumption expenditure data is hard to obtain, and there is a lack of open-source estimation methods and comparative modeling studies on both the supply and demand sides of consumption. This study proposes a framework for estimating offline consumption expenditure using review big data, which is simple and effective. Using store density as the supply-side proxy and estimated offline expenditure as the demand-side proxy, machine learning models are employed to analyze the spatial distribution of consumer geography and its nonlinear links with urban environmental factors. Results show that the review big data-based estimation accurately reflects actual expenditure patterns, solving the data accessibility issue. Store density and consumer expenditure exhibit a consistent spatial pattern, with higher thresholds needed to boost expenditure than to increase store density. These findings offer valuable insights for urban planners and businesses, emphasizing the importance of considering nonlinear impacts and threshold effects on store layout. This study provides a reliable, open-source method for estimating offline consumption expenditure, advancing research and practice in retail and consumer geography, and enriching data sources and modeling perspectives in economic geography.