Crowdsourcing-data-based dynamic measures of accessibility to business establishments and individual destination choices

摘要

Characterizing accessibility is essential in research efforts to investigate the reciprocal impacts between human activities and urban space. Although previous studies have proposed various accessibility indicators, they are mostly static in nature, capturing only the spatial dimension of accessibility. The advancement of the research agenda in activity-based studies calls for improved accessibility measures that can capture urban dynamics and human activity-travel patterns at more fine-grained resolutions, both spatially and temporally. In this paper, we draw on the rich spatial and temporal information from online crowdsourcing data on points-of-interest and propose a set of new measures of accessibility to opportunities and services that incorporate the dynamics of businesses environment in Singapore. The new measures consider both the spatial distribution of business establishments and their open time over the day. We then feed the new and traditional accessibility measures into destination choice models to assess whether capturing the temporal variation in accessibility and business environment help to improve the model performance. We find that the model with dynamic accessibility measures outperforms models with traditional static accessibility measures. The results suggest that travelers are likely to consider both the spatial and temporal dimensions of accessibility when choosing destinations for their activities. The proposed dynamic accessibility measures contribute to the research and planning for accessibility by capturing the temporal dimension of accessibility that are not sufficiently addressed in traditional urban built environment measures.