Research Article

Spatial Variation of Taxi Demand Using GPS Trajectories and POI Data

Table 5

Comparison of findings with existing studies.

StudyCityRidershipModelSignificant factors

This studyQingdao, ChinaTaxiGWRResidential density (+), housing price (+), road density (+), parking lot density (+), bus station density (+), residential area (), commercial area (+), public service area (), other land use (), land use mix ()
Tu et al. (2018)Shenzhen, ChinaTaxiGWREmployment (), income (), land use mix (), road density (+), bus accessibility (), metro accessibility (+)
Yang et al. (2018)Washington, DC, USATaxiOLS regression modelBlock (+), metro (+), bus (−), airport (+), pop (+), ResiDens (), RetaiDens (+), OffDens (), IndDens (+), OthDens (+), EmpEntropy ()
Qian and Ukkusuri (2015)New York, USATaxiGWRCommuting time (−), highly educated population (+), median income (), road density (+), subway accessibility (+), commercial area ()

+positive effect, − negative effect, negative and positive effects.