Inferring transit-based health seeking patterns from smart card data - A case study in Beijing, China

Massive electronic trip records have recently been utilized to infer people's trips for healthcare. Many inferential methods were developed to derive healthcare trips by taxi using GPS trajectory records, but little attention is paid to public transit, as a common travel mode for healthcare. This paper proposes a method to fill this gap by mining a big data of smart transit cards with spatio-temporal constraints. We demonstrate and validate this method in Beijing, China. The inferred trips achieve a high degree of consistency, in space and time, with empirically observed trips from a survey. The inferred trips are further used to identify spatial disparities in transit-based access to healthcare, which might have been overlooked by health policy makers.