Order allocation strategy for online car-hailing platform in the context of multi-party interests

Abstract

As an emerging internet service, online car-hailing allows users to do the car-hailing via their mobile phone instead of calling cars on the road. Though the convenient service provided by online car-hailing platforms, the following problems encountered in the process of real-time order-matching still cannot be well addressed: (1) the imbalance of the accessible cars in different areas, and (2) the conflict of interests among the drivers, platforms, and passengers. To address those problems, this paper proposes a set of order allocation strategy that fully consider the changing regional characteristics and the interests of multiple parties. Firstly, a dynamic region division method is proposed to identify the corresponding area types for real-time car-hailing. Then, a novel order allocation strategy with the regional multi-party subsidy mechanism is developed for different types of areas. Lastly, to help obtain a better performance for multiple parties, a parameter optimization method for the multi-party subsidy mechanism is introduced. The proposed order allocation strategy is evaluated on the real-world dataset sourced from an online car-hailing company. Our experiments verify the feasibility and effectiveness of the proposed method, without reducing drivers’ benefits, the order response time and order response rate are greatly improved, and the cost in all is acceptable for passengers and the platform. It is illustrated that a suitable regional multi-party subsidy mechanism could help online car-hailing platforms obtain higher passenger satisfaction than that without a regional multi-party subsidy mechanism.