A prediction method and device for task matching

Abstract

An embodiment of the present invention provides a method and device for predicting task matching, wherein the method includes: obtaining task data browsed by candidate task performers as a sample set, wherein each task data contains at least two attributes; The tree determines the priority attributes of the candidate task executors; when a new task arrives, the prediction result of whether the candidate task executor will accept the new task is obtained according to the Bayesian model. The present invention uses big data analysis means to draw portraits for candidate task performers (such as drivers), and predict whether the candidate task performers may accept new tasks from a more comprehensive perspective. In this way, new tasks can be more targeted when pushed to candidate task executors, ensuring the precise matching between order tasks and candidate task executors, and realizing accurate push of order tasks, thereby improving the delivery efficiency of the entire logistics platform and reducing logistics costs. In addition, there are beneficial effects such as reducing the time consumption of screening tasks.

Description

technical field [0001] The embodiments of the present invention relate to the technical fields of the Internet and logistics, and in particular to a prediction method and device for task matching. Background technique [0002] With the rapid development of the national economy and Internet technology, my country's current logistics scale has shown geometric growth. According to statistics, the total social logistics in 2017 has reached a scale of 252.8 trillion yuan. Although many logistics network platforms and freight companies have emerged, the slow improvement of logistics distribution efficiency has not been significantly improved, and logistics costs are still high. [0003] There are various complex links in logistics distribution. In order to reduce costs and improve logistics distribution efficiency, existing technologies focus on how to reduce the waiting time in the distribution link, reduce the empty load rate of trucks, and predict the number of orders to prepar...

AI Technical Summary

Problems solved by technology

[0004] However, the inventor found in the process of realizing the present invention that although the existing logistics technical solutions optimize the logistics distribution process by adding more logistics-related factors to the logistics scheduling model, and even predict the logistics status at the next moment, the logistics distribution The improvement of efficiency is still limited, and no breakthrough has been made

The reason is that the existing technology still does not achieve the precise matching of the two resources of the driver and the order task. More in-depth speaking, the various logistics models constructed by the existing technology equate "vehicle" with "people". ", only saw the vehicle, but ignored the existence of the driver, did not consider or integrate the individual factors of the driver

Drivers are valuable resources in the logistics industry and are the executors of the "last mile". Therefore, drivers as a group directly affect the delivery efficiency at the end of logistics. For example, whether a driver has a certain tendency towards order tasks, whether his experience can Competent for order tasks, which drivers are more willing to accept which orders, whether orders can be pushed to more suitable drivers in a more targeted manner, etc., but when the existing technology matches the two resources of drivers and order tasks, in fact, It only achieved the matching of vehicles and order tasks, ignoring the existence of drivers, which eventually led to bottlenecks in optimizing logistics delivery efficiency

Images