Using Beijing as a case study, the correlation between accessibility and commercial rent is initially explored in this research by analyzing leasing data sourced from the Anjuke platform. This study aims to refine the theory of rent within the Chinese urban context to distinguish between favorable and unfavorable locations. Subsequently, POI data from leisure consumption shops combined with online reviews extracted from Dazhong Dianping in 2015 and 2019 are used to scrutinize the development trend of urban leisure consumption spaces at both the macro (city) and micro (building) levels. For macro-scale accessibility, integration and choice measures of road networks are employed via the space syntax tool to assess proximity to urban centers and the road hierarchy. Regarding micro-scale accessibility, we develop methods to compute proximity to streets and identify ground- floor shops, enabling the measurement of the proximity to ground-floor storefronts from both horizontal and vertical perspectives. Based on the macro and micro accessibility calculations for each shop, we visualize their overall changes in 2015 and 2019, compare the differences between shops and online reviews of shops, and examine the robustness of the results for different types of leisure consumption shops (i.e., entertainment, sports, retail, catering, and personal care).
The research outcomes indicate 1) a trend toward low-accessibility areas for both physical stores and electronic word- of-mouth, revealing the rise of hidden space across all scales; 2) that electronic word-of-mouth, as represented by the number of online reviews, tends to concentrate in advantageous locations at the macroscale while penetrating to hidden locations at the microscale, thus reflecting the dual effect of EWoM; 3) that there are weaker changes at the macroscale than at the microscale, suggesting that the development of leisure consumption space in Beijing during this period primarily shifted toward disadvantaged areas at the microscale; and 4) that the study validates the robustness of the results by comparing different types of leisure consumption spaces. The study findings can enrich the understanding of the location choices of urban leisure consumption spaces in the digital era and facilitate the refinement of urban renewal and management at a fine scale.