Granted Patent B2

US 12,192,847 · App. 17/660,408 · Granted Jan 7, 2025

Inventors: Yichen Zhao (Beijing, CN); Qian Liu (Beijing, CN); Chao Shen (Beijing, CN); Minghui He (Beijing, CN); Jiantao Yang (Beijing, CN); Weihuan Shu (Beijing, CN); Li Ma (Beijing, CN)

Assignee: BEIJING DIDI INFINITY TECHNOLOGY AND DEVELOPMENT CO., LTD.

H04W4/021G01C21/3885G06F16/24578H04W4/025

Abstract

Claims (89)

1. A system for recommending a pick-up point (PUP), comprising:

at least one storage medium including a set of instructions;

at least one processor in communication with the at least one storage medium, wherein executing the set of instructions, the at least one processor is directed to perform operations including:

obtaining information of a searched point of interest (POI) determined by a user;

determining a first set of candidate PUPs based at least on the information of the searched POI, the first set of candidate PUPs including a plurality of candidate PUPs;

determining at least one recommended PUP through a preset algorithm based at least on the first set of candidate PUPs, and displaying the at least one recommended PUP to the user;

wherein the determining at least one recommended PUP through a preset algorithm based at least on the first set of candidate PUPs includes:

determining a set of security PUPs based on the first set of candidate PUPs, wherein the set of security PUPs includes a PUP with a small parking risk to avoid a driver from picking up a passenger in the parking violation road section; and

determining the at least one recommended PUP through the preset algorithm based on the set of security PUPs.

2. The system of claim 1 , wherein the searched POI includes one or more points of interest (POIs) selected by the user based on a set of recommended POIs; and

the determining a first set of candidate PUPs based at least on the information of the searched POI includes:

obtaining a positioning location of the user and historical behavior information of the user;

determining a set of candidate POIs based on the positioning location of the user and the historical behavior information of the user, the set of candidate POIs including a plurality of candidate POIs;

determining the set of recommended POIs based on the set of candidate POIs, the set of candidate POIs including at least one recommended POI; and

determining the first set of candidate PUPs based at least on the information of the searched POI selected by the user from the set of recommended POIs.

3. The system of claim 2 , wherein the determining a set of candidate POIs based on the positioning location of the user and the historical behavior information of the user includes:

determining a set of first POIs including one or more first POIs within a preset range of the positioning location, a spatial distance between each of the one or more first POIs with the positioning location of the user being less than a first preset distance threshold;

determining, based on the historical behavior information of the user, a set of second POIs including one or more second POIs that are historically clicked and/or searched by the user; and

determining the set of candidate POIs based on the set of first POIs and the set of second POIs.

4. The system of claim 2 , wherein the determining the set of recommended POIs based on the set of candidate POIs includes:

obtaining first feature information of the plurality of candidate POIs;

obtaining scoring values of the plurality of candidate POIs by scoring the plurality of candidate POIs based on the first feature information;

obtaining a screened set of candidate POIs by screening the plurality of candidate POIs based on similarity degrees between the plurality of candidate POIs; and

determining the set of recommended POIs based on a scoring value of each of one or more candidate POIs in the screened set of candidate POIs.

5. The system of claim 4 , wherein the first feature information of a candidate POI includes at least one of a self-attribute feature of the candidate POI, a relationship feature between the candidate POI and the positioning position of the user, or a feature of a historical click rate of the candidate POI; wherein

the self-attribute feature of the candidate POI includes at least one of a type of the candidate POI, a count of candidate PUPs recalled by the candidate POI, or a hotness of the candidate POI; and

the relationship feature between the candidate POI and the positioning location of the user includes at least one of a distance feature or a feature whether to cross a road.

6. The system of claim 4 , wherein the determining the set of recommended POIs based on a scoring value of each of one or more candidate POIs in the screened set of candidate POIs includes:

obtaining a sorting result by sorting the one or more candidate POIs in the screened set of candidate POIs based on the scoring value of each of the one or more candidate POIs; and

determining the set of recommended POIs based on the sorting result.

7. The system of claim 1 , wherein

the set of security PUPs including at least one security candidate PUP.

8. The system of claim 1 , wherein the determining a set of security PUPs based on the first set of candidate PUPs includes:

obtaining information of one or more parking violation road sections;

determining, based on the information of the one or more parking violation road sections, whether each of the plurality of candidate PUPs in the first set of candidate PUPs meets a preset security condition; and

adding one or more candidate PUPs each of which meets the preset security condition to the set of security PUPs.

9. The system of claim 8 , wherein the preset security condition includes a condition that a parking risk value of the candidate PUP is less than a preset risk threshold; wherein

the parking risk value is a ratio of a count of tickets in orders where the candidate PUP is a starting point to a total count of the orders within a first preset time threshold.

10. The system of claim 1 , wherein the determining the at least one recommended PUP through the preset algorithm based on the set of security PUPs includes:

obtaining second feature information of the at least one candidate PUPs in the set of security PUPs; and

determining the at least one recommended PUP through the preset algorithm based on the second feature information.

11. The system of claim 1 , wherein the preset algorithm includes a trained PUP recommendation model.

12. The method of claim 11 , wherein the trained PUP recommendation model is obtained according to a following process including:

obtaining data of a plurality of historical orders;

obtaining, from the data of the plurality of historical orders, a billing point and a real PUP of each historical order in the plurality of historical orders;

screening the data of the plurality of historical orders based on a spatial position relationship between the billing point and the real PUP of each historical order;

designating the screened data as training samples; and

obtaining the trained PUP recommendation model by training a model based on the training samples.

13. The system of claim 12 , wherein the obtaining a billing point and a real PUP of each historical order from the data of the plurality of historical orders includes:

obtaining a motion trajectory of a service provider terminal and a motion trajectory of a service requester terminal in each historical order;

determining, based on the motion trajectory of the service provider terminal, at least one stopping point of the service provider terminal;

determining, based on the motion trajectory of the service requester terminal, at least one stopping point of the service requester terminal; and

determining the real PUP based on the at least one stopping point of the service provider terminal and the at least one stopping point of the service requester terminal.

14. The system of claim 13 , wherein the determining, based on the motion trajectory of the service requester terminal, at least one stopping point of the service requester terminal includes:

determining, based on the motion trajectory of the service requester terminal, whether a dwell time of the service requester terminal at a motion trajectory point is larger than a second preset time threshold;

in response to that the dwell time of the service requester terminal at the motion trajectory point is larger than the second preset time threshold, determining whether motion trajectory points within the second preset time threshold are available to be clustered; and

in response to that the motion trajectory points within the second preset time threshold are available to be clustered, determining the at least one stopping point of the service requester terminal based on the clustered motion trajectory points of the service requester terminal.

15. The system of claim 14 , wherein the determining the real PUP based on the at least one stopping point of the service provider terminal and the at least one stopping point of the service requester terminal includes:

obtaining a plurality of intersection points between the motion trajectory points of the service provider terminal and motion trajectory points of the service requester terminal;

determining a first distance between one intersection point among the plurality of intersection points and the at least one stopping point of the service provider terminal, and determining a second distance between the intersection point and the at least one stopping point of the service requester terminal; and

determining an intersection point of the plurality of intersection points to be the real PUP, the first distance and the second distance corresponding to the intersection point being minimum.

16. The system of claim 12 , wherein the screening the data of the plurality of historical orders based on a spatial position relationship between the billing point and the real PUP in each historical order includes:

determining a distance between the billing point and the real PUP in each historical order;

in response to the distance between the billing point and the real PUP in the historical order is less than or equal to a second preset distance threshold, determining whether the billing point and the real PUP in the historical order are located in a same road section; and

in response to that the billing point and the real PUP in the historical order are located in the same road section, determining that the historical order passes the screening.

17. The system of claim 12 , wherein the designating the screened data as training samples; and obtaining the trained PUP recommendation model by training a model based on the training samples includes:

for each historical order that passes the screening,

obtaining a searched POI and a PUP in the historical order;

determining a feature of the searched POI and a feature of the PUP in the historical order;

determining a label of the historical order based on a spatial position relationship between a recommended PUP and the real PUP of the historical order; and

obtaining the trained PUP recommendation model by training, based on the feature of the searched POI and the feature of the PUP in the historical order and the label of the historical order, an initial machine learning model.

18. The system of claim 1 , wherein the determining at least one recommended PUP through a preset algorithm based at least on the first set of candidate PUPs includes:

determining second feature information of at least one candidate PUP in the first set of candidate PUPs;

determining a scoring value of each candidate PUP by scoring each candidate PUP in the at least one candidate PUP through the preset algorithm based at least on the second feature information of the at least one candidate PUP; and

determining the at least one recommended PUP based on the scoring value of each candidate PUP in the at least one candidate PUP.

19. A method for recommending a pick-up point (PUP) implemented by a computer device including at least one processor and a storage device, comprising:

obtaining information of a searched point of interest (POI) determined by a user;

determining a first set of candidate PUPs based at least on the information of the searched POI, the first set of candidate PUPs including a plurality of candidate PUPs;

determining at least one recommended PUP through a preset algorithm based at least on the first set of candidate PUPs, and displaying the at least one recommended PUP to the user;

wherein the determining at least one recommended PUP through a preset algorithm based at least on the first set of candidate PUPs includes:

determining a set of security PUPs based on the first set of candidate PUPs, wherein the set of security PUPs includes a PUP with a small parking risk to avoid a driver from picking up a passenger in the parking violation road section; and

determining the at least one recommended PUP through the preset algorithm based on the set of security PUPs.

20. A non-transitory computer readable storage medium, comprising computer instructions that, when executed by at least one processor, direct the at least one processor to perform methods including:

obtaining information of a searched point of interest (POI) determined by a user;

determining a first set of candidate PUPs based at least on the information of the searched POI, the first set of candidate PUPs including a plurality of candidate PUPs;

determining at least one recommended PUP through a preset algorithm based at least on the first set of candidate PUPs, and displaying the at least one recommended PUP to the user;

wherein the determining at least one recommended PUP through a preset algorithm based at least on the first set of candidate PUPs includes:

determining a set of security PUPs based on the first set of candidate PUPS, wherein the set of security PUPs includes a PUP with a small parking risk to avoid a driver from picking up a passenger in the parking violation road section; and

determining the at least one recommended PUP through the preset algorithm based on the set of security PUPs.

Assignments (3)

Priority Claims (4)

Continuity (2)

References Cited (39)