In the real world, some taxi drivers are more capable of finding new passengers and make more money than others. It is attractive to learn such service skills to earn more for low-performance taxi drivers. Thanks to the widely available large-scale taxi GPS trajectory data, service strategies of taxi drivers that are hidden in their trajectory data can be uncovered by data mining techniques. In this chapter, we investigate taxi service strategies from three perspectives, i.e., passenger-hunting, passenger-sending, and service-area preference. We build a feature matrix to link extracted taxi service strategies from GPS trajectory data to the resulted taxi driver’s revenue in fine granularity. We can inform which strategies are efficient or inefficient by evaluating the correlation values between service strategies and the resulted revenue. A positively higher correlation value implies a corresponding more efficient strategy. To demonstrate that the extracted taxi service strategies with our proposed approach well characterize the driving behavior and performance of taxi drivers, we build a regression model and predict the revenue of taxi drivers based on their strategies, achieving a prediction residual as less as 2.35 RMB/h.
Hunting or Waiting: Earning More by Understanding Taxi Service Strategies
Enabling Smart Urban Services with GPS Trajectory Data ; Chapter : 4 ; 71-94
2021-04-02
24 pages
Article/Chapter (Book)
Electronic Resource
English
Understanding Taxi Service Strategies From Taxi GPS Traces
Online Contents | 2015
|Taxi-RS: Taxi-Hunting Recommendation System Based on Taxi GPS Data
Online Contents | 2015
|Online Contents | 1995
|