A Map-Matching Service Designed for Courier Trajectories

In express delivery, couriers will generate a mass of trajectory logs when delivering shipments. To analyze these logs is of great value for the promotion of express delivery service. For any research based on trajectory data, map-matching plays an important role, so in this article, we design a map-matching service specially for courier trajectories. As far as we know, existing map-matching algorithms are designed mainly for cars or walks, or ignoring means of transportation. Although these methods can be applied to courier trajectory map-matching, the accuracy of them can hardly be guaranteed as they ignore the characteristics of courier trajectories. To solve this problem, we design a new map-matching service based on a map-matching algorithm called Courier Trajectory Based Map-Matching (CTB-Matching), which is specially used to deal with courier trajectories. Courier trajectories have some characteristics different from traditional trajectories. Firstly, couriers have to deliver shipments at different sites, so the trajectories seem more irregular, which is called as "fragmentation" problem. Secondly, unlike cars, couriers' positioning information is mainly generated by Wi-Fi location system, which is not precise as Global Position System (GPS), so the location deviation problem should be taken into account. What's more, couriers usually use electric bikes for delivery, which travel slower than cars, and are less likely to be influenced by traffic. Therefore, the speed or temporal analysis for cars is not suitable here. Based on the analysis of current algorithms and the problems stated above, this paper designs a map-matching service for courier trajectory data. The experiments verify that our service performs better when dealing with courier trajectory data. Besides, our service is efficient with low time complexity and space complexity, making our service responses with low latency.