CN113780955A

CN113780955A - Position checking method and device, electronic equipment and storage medium

Info

Publication number: CN113780955A
Application number: CN202111095250.4A
Authority: CN
Inventors: 丁一; 丁凡; 沈国斌; 江东哲; 何田
Original assignee: Rajax Network Technology Co Ltd
Current assignee: Rajax Network Technology Co Ltd
Priority date: 2021-09-17
Filing date: 2021-09-17
Publication date: 2021-12-10

Abstract

One or more embodiments of the present specification provide a position verification method and apparatus, an electronic device, and a storage medium; the method can comprise the following steps: determining a place to be verified and a plurality of reference places corresponding to the place to be verified, and acquiring declaration positions of the place to be verified and the reference places; determining an estimated travel distance between the to-be-verified place and each reference place according to travel related information between the to-be-verified place and each reference place, wherein the travel related information comprises travel time information between the to-be-verified place and each reference place; and positioning the place to be verified according to the declaration position of each reference place and the estimated travel distance to obtain the estimated position of the place to be verified, matching the estimated position with the declaration position of the place to be verified, and judging whether the declaration position of the place to be verified passes verification or not according to the matching result.

Description

Position checking method and device, electronic equipment and storage medium

Technical Field

One or more embodiments of the present disclosure relate to the field of positioning technologies, and in particular, to a position verification method and apparatus, an electronic device, and a storage medium.

Background

In the related art, the physical location of a certain location may be uploaded to a specific platform to complete registration. For example, in an application scenario of instant delivery, for a delivery platform, an accurate physical location of a merchant is crucial to order scheduling, and reasonable order scheduling can ensure that delivery of a delivery order is completed before a promised time, so as to avoid timeout.

Therefore, accurate registration of the physical location is an important link. However, in the registration process, the registered physical location is often deviated due to human factors. For example, merchant locations are registered by the merchant itself, and some erroneous merchant locations may be due to human error or merchant location updates. But still a part of the errors are due to merchant location fraud. For example, a merchant registered at a shopping mall often means better service and higher food quality than a residential area. Thus, some merchants may intentionally register the wrong location (e.g., in a mall) to attract more online consumers that do not have to be physically present, but rather only need to be delivered by the rider, i.e., a location fraud problem.

Disclosure of Invention

In view of this, one or more embodiments of the present disclosure provide a position verification method and apparatus, an electronic device, and a storage medium.

To achieve the above object, one or more embodiments of the present disclosure provide the following technical solutions:

according to a first aspect of one or more embodiments of the present specification, there is provided a position verification method, including:

determining a place to be verified and a plurality of reference places corresponding to the place to be verified, and acquiring declaration positions of the place to be verified and the reference places;

determining an estimated travel distance between the to-be-verified place and each reference place according to travel related information between the to-be-verified place and each reference place, wherein the travel related information comprises travel time information between the to-be-verified place and each reference place;

and positioning the place to be verified according to the declaration position of each reference place and the estimated travel distance to obtain the estimated position of the place to be verified, matching the estimated position with the declaration position of the place to be verified, and judging whether the declaration position of the place to be verified passes verification or not according to the matching result.

Optionally, the determining a location to be verified and a plurality of reference locations corresponding to the location to be verified includes:

acquiring a place set comprising a plurality of places located within a preset range;

and traversing each place in the place set to sequentially take the current place as the place to be checked and take other places different from the current place as the reference places.

Optionally, the determining, according to the travel related information between the location to be verified and each reference location, an estimated travel distance between the location to be verified and each reference location includes:

inputting the travel related information into a distance model, wherein the distance model is obtained by training historical travel related information of a plurality of historical travels by adopting a supervised learning algorithm, and the historical travel related information is marked with standard travel distances of the historical travels;

and determining the estimated travel distance between the to-be-checked place and each reference place according to the output result of the distance model.

Optionally, in a case that the location to be verified and the reference point are physical stores, the method further includes:

acquiring the store leaving time information of any distribution order accepted by a distributor and the store arriving time information of the next distribution order;

and determining travel time information between a first physical store corresponding to the any delivery order and a second physical store corresponding to the next delivery order according to the departure time information and the arrival time information.

Optionally, the positioning the location to be verified according to the declared position of each reference location and the estimated travel distance to obtain the estimated position of the location to be verified includes:

searching a target place, wherein errors between the standard travel distance between the position of the target place and the statement positions of the reference places and the estimated travel distance meet preset error conditions;

taking the position of the target location as the estimated position.

Optionally, the target location is searched in the following manner:

representing said target location, q representing an arbitrary location, q^*Representing the place to be checked;

P＝{p_i,i＝1...N}，p_irepresenting a reference location i, N representing the number of said reference locations;

d_s(p_iq) represents p_iThe standard trip distance between the declared location of (a) and the location of q;

d_e(p_i,q^*) Represents p_iAnd q is^*The estimated travel distance therebetween.

Optionally, the standard trip distance includes a trip distance between the position of the target location calculated using the map data and the declared position of each reference location.

Optionally, in a case that the location to be verified and the reference point are physical stores, the travel related information between the location to be verified and any physical store further includes at least one of the following:

the number of delivery orders contained in the area to which the to-be-verified place belongs, the number of delivery orders contained in the area to which any one of the physical stores belongs, road condition information between the to-be-verified place and any one of the physical stores, current delivery state information of a delivery party who goes to another place from one of the to-be-verified place and any one of the physical stores, and delivery preference information of the delivery party.

According to a second aspect of one or more embodiments of the present specification, there is provided a position verification apparatus including:

the verification device comprises a determination unit, a verification unit and a verification unit, wherein the determination unit determines a place to be verified and a plurality of reference places corresponding to the place to be verified, and acquires declaration positions of the place to be verified and the reference places;

the pre-estimation unit is used for determining the estimated travel distance between the to-be-verified place and each reference place according to travel related information between the to-be-verified place and each reference place, wherein the travel related information comprises travel time information between the to-be-verified place and each reference place;

and the verification unit is used for positioning the place to be verified according to the declaration position of each reference place and the estimated travel distance to obtain the estimated position of the place to be verified, matching the estimated position with the declaration position of the place to be verified, and judging whether the declaration position of the place to be verified passes verification or not according to a matching result.

Optionally, the determining unit is specifically configured to:

Optionally, the estimating unit is specifically configured to:

Optionally, the estimating unit is further configured to:

Optionally, the verification unit is specifically configured to:

taking the position of the target location as the estimated position.

Optionally, the target location is searched in the following manner:

According to a third aspect of one or more embodiments of the present specification, there is provided an electronic apparatus including:

a processor;

a memory for storing processor-executable instructions;

wherein the processor implements the position verification method as described in any of the above embodiments by executing the executable instructions.

According to a fourth aspect of embodiments of the present disclosure, there is provided a computer-readable storage medium having stored thereon computer instructions which, when executed by a processor, implement the steps of the method as in any one of the above-mentioned embodiments.

In the technical solution of the present specification, a place to be verified has a declaration position, that is, a physical position declared for the place to be verified. It should be noted that the physical location is not necessarily accurate, and may deviate from the actual physical location. Therefore, a plurality of reference places are obtained to check the declaration position of the place to be checked by using the declaration positions of the reference places.

Specifically, the travel time between two locations is correlated with the travel distance between the two locations, and based on the correlation between the travel time and the travel distance, the travel distance between any two locations can be estimated using the travel time. Then, the travel distances between the location to be verified and the reference locations may be estimated to obtain estimated travel distances, and the obtained estimated travel distances are used as actual travel distances (understood as real travel distances) between the location to be verified and the reference locations, so that the location to be verified is located according to the declared location and the actual travel distances of the reference locations to obtain an estimated location (at this time, the estimated location is considered as an estimated real physical location). It will be appreciated that if the declared location of the site to be verified has a small error from the actual physical location, the declared location should match the estimated location. Thus, whether the claim location passes the check may be based on whether the estimated location matches the claim location. Through the mode of verifying the physical position, manual work can be avoided going to the site for verification, and therefore cost is reduced on the premise of ensuring verification accuracy.

Drawings

Fig. 1 is a flowchart of a location verification method according to an exemplary embodiment.

FIG. 2 is a schematic diagram of a store location error provided by an exemplary embodiment.

FIG. 3 is a schematic diagram of combined time null data provided by an exemplary embodiment.

FIG. 4 is a graphical depiction of a distance traveled by a map provided by an exemplary embodiment.

Fig. 5 is a schematic structural diagram of an apparatus according to an exemplary embodiment.

Fig. 6 is a block diagram of a position checking apparatus according to an exemplary embodiment.

Detailed Description

Reference will now be made in detail to the exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, like numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with one or more embodiments of the present specification. Rather, they are merely examples of apparatus and methods consistent with certain aspects of one or more embodiments of the specification, as detailed in the claims which follow.

It should be noted that: in other embodiments, the steps of the corresponding methods are not necessarily performed in the order shown and described herein. In some other embodiments, the method may include more or fewer steps than those described herein. Moreover, a single step described in this specification may be broken down into multiple steps for description in other embodiments; multiple steps described in this specification may be combined into a single step in other embodiments.

Taking an application scenario of online instant delivery as an example, online orders (such as food, medicine, etc.) are delivered to users from physical stores (such as restaurants, drug stores, supermarkets, etc.) of merchants within a short time (such as 30 minutes) by a deliverer (such as a rider). If the time that the user receives the goods exceeds the time limit, the platform will compensate the user. In such a platform, the actual physical location of the brick and mortar store is critical to the allocation of the order to the most appropriate rider, i.e., order scheduling. Reasonable order scheduling can guarantee that an order completes delivery before the time of commitment to avoid timeouts. In fact, the locations of these stores are registered by the merchant themselves at the distribution platform and are therefore not all their exact physical locations. Some erroneous location records may be due to human error or merchant location updates, but also in part due to deliberate manipulation, i.e., location fraud issues.

The above-mentioned problem of location error will seriously affect the business of the online instant delivery platform, such as causing problems of improper order distribution, delivery route deviation and unnecessary overtime compensation for the user. Therefore, the distribution platform needs to identify those merchants with the wrong store location and find their true locations.

In the related art, the delivery platform may require the rider to report the wrong place by the way when delivering, but the manner of the rider reporting increases the burden on the rider and the labor cost is high. In contrast, the present specification provides a position verification scheme that can reduce costs while ensuring verification accuracy.

Referring to fig. 1, fig. 1 is a flowchart illustrating a location verification method according to an exemplary embodiment. As shown in fig. 1, the method may include the steps of:

step 102, determining a place to be verified and a plurality of reference places corresponding to the place to be verified, and acquiring declaration positions of the place to be verified and the reference places.

The declared location of a place to be verified, i.e., the physical location declared for the place to be verified. It should be noted that the physical location is not necessarily accurate, and may deviate from the actual physical location. In a real business scenario, the physical location registered or registered for a certain place is not necessarily accurate. Taking the online instant distribution scenario as an example, the physical location of the merchant physical store is manually registered by the merchant (for example, adding a text address, uploading a GPS location of the mobile device, or manually selecting a point on a map), and then converted into GPS coordinates by the system. Therefore, the positions of the merchant brick-and-mortar stores are easily deviated due to manual operation or manual counterfeiting.

By way of example, FIG. 2 shows an example of a merchant with a critical mis-location. As shown in fig. 2, solid dots represent registered locations of the merchant a store, and open dots represent actual locations of the merchant a store. When a new delivery order appears at the merchant a, it is assumed that two riders exist on the delivery platform and are respectively located at the registered positions of the merchant B and the merchant C. Then, in the event of a mis-location of merchant a, the delivery platform may mis-assign the delivery order to the rider located at merchant C, while the rider located at merchant B is actually closer to the location of merchant a's store.

In contrast, the position verification scheme of the present specification performs cross validation on a plurality of locations by using a travel distance between two locations to verify whether an error exists in the corresponding location. It should be noted that the distance used in this specification is travel distance (travel distance), i.e., distance, and not displacement. For example, if a rider distributes on foot, then the distance traveled is the rider's walking distance; if the rider is using an electric bicycle for distribution, the distance traveled is the riding distance of the rider. The principle of cross-validation is illustrated below in conjunction with fig. 2.

And cross-verifying whether the registered position of the store position of the merchant is suspicious or not by using the travel distance from one store to other nearby stores. As shown in fig. 2, it is assumed that the registered locations of the stores of the merchant a are wrong, and the registered locations of the stores of the merchant B and the merchant C are accurate. Then, from A, B, C the registered locations of the three merchant stores and the travel distance between each merchant store, it can be found that: (1) the actual travel distance between a and B is much shorter than the travel distance calculated from the registered position of a; (2) the actual travel distance between a and C is much longer than the travel distance calculated from the registered position of a; (3) the actual travel distance between B and C is close to the travel distance calculated from the registered positions of B and C. Therefore, in practical application, the above-mentioned manner of cross-verifying the travel distance may be adopted, and if the deviation condition on the travel distance occurs, it may be deduced reversely that there is an error in the registered location of the merchant a store.

Based on the above-mentioned principle of cross-validation, in order to further improve the accuracy of validation, the distance between the mutually validated sites should be avoided too far, because a long distance of travel may cause large variation noise to the data. In order to solve the above problem, a smaller range may be preset, and the sites to be verified included in the range are drawn into the site set to sequentially perform the cross-validation step for each site in the set. In other words, each place in the place set can be traversed, and the cross-validation step is performed once for each place in the place set, and in each cross-validation step, the place currently polled is the place to be checked, and the other places are reference places. The preset range is controlled within a proper value (for example, within 200m, 500m, 800m, and the like), so that the problem of high noise can be avoided, that is, the location to be verified and other nearby locations can mutually verify the correctness of the location, and the positioning is facilitated. Based on the setting of the preset range, in the process of determining the location to be verified and a plurality of reference locations corresponding to the location to be verified, a location set including a plurality of locations within the preset range can be obtained, and then each location in the location set is traversed, so that the current location is sequentially used as the location to be verified, and other locations different from the current location are used as reference locations. Of course, the specific value of the preset range can be flexibly set according to the actual situation, and the specification does not limit this.

Or, besides the above-mentioned way of dividing the location set to traverse each location in the set for cross validation, the reference location may not be the declared location, and may also be the actual physical location of the reference location (such as a manual site survey, or a location that passes verification), so that it is only necessary to verify the declared location of the location to be verified by using the actual physical material of the reference location, and the above-mentioned operation of traversal need not be implemented.

And 104, determining the estimated travel distance between the to-be-verified place and each reference place according to travel related information between the to-be-verified place and each reference place, wherein the travel related information comprises travel time information between the to-be-verified place and each reference place.

As can be seen from the principle of the cross-validation, in order to ensure the accuracy of comparing the travel distances in the cross-validation process, the actual travel distance between two stores should be made accurate enough. While at the same time avoiding manual work to field survey the actual travel distance between two stores, a means of estimating the actual travel distance (hereinafter referred to as an estimated travel distance) may be employed to reduce costs. Then, the accuracy of estimating the travel distance should be improved as much as possible.

As an exemplary embodiment, the actual trip distance may be estimated using the correlation of the trip related information and the trip distance. In one case, the trip-related information includes trip time information. In particular, the travel time between two locations is related to the travel distance between the two locations. For example, in the case of the same vehicle, the longer the travel distance between two points, the longer the travel time is consumed, i.e., the travel time is positively correlated with the travel distance.

Based on the correlation between travel time and travel distance, travel time may be used to estimate travel distance between any two locations. Then, the travel distances between the location to be verified and the reference locations may be estimated to obtain estimated travel distances, and the obtained estimated travel distances are used as actual travel distances (understood as real travel distances) between the location to be verified and the reference locations, so that the location to be verified is located according to the declared location and the actual travel distances of the reference locations to obtain an estimated location (at this time, the estimated location is considered as an estimated real physical location). It will be appreciated that if the declared location of the site to be verified has a small error from the actual physical location, the declared location should match the estimated location. Thus, whether the claim location passes the check may be based on whether the estimated location matches the claim location. Through the mode of verifying the physical position, manual work can be avoided going to the site for verification, and therefore cost is reduced on the premise of ensuring verification accuracy.

Wherein, a supervised learning algorithm can be adopted to learn the correlation between the travel time and the travel distance so as to train and obtain a distance model for estimating the actual travel distance. Aiming at the training process of the supervised learning algorithm, input sample data is called a training set, the sample data in the training set has a definite identification or result (namely a sample label), when the supervised learning algorithm is used for establishing a distance model, the supervised learning algorithm establishes a learning process, a prediction result is compared with an actual result of the training set, and the prediction model is continuously adjusted until the prediction result of the model reaches an expected accuracy rate. For example, common supervised learning algorithms include logistic regression, neural networks, decision trees, support vector machines, bayesian classifiers, and the like.

Therefore, historical trip related information (such as the trip time information) of a plurality of historical trips can be acquired as sample data, and meanwhile, the historical trip related information is labeled (namely, the sample data is marked), and the corresponding label is a standard trip distance of the acquired historical trips. In order to make the prediction accuracy of the trained distance model higher, the accuracy of the label should be ensured. For this, the travel distance between two points involved in the historical travel may be calculated using map data (the position information in the map data is more accurate, has passed verification) as the standard travel distance. Alternatively, the standard trip distance may also be derived from a manual field survey. Of course, any other method can be adopted as long as the standard travel distance can accurately represent the actual travel distance.

Based on the training of the distance model, for the determination of the estimated travel distance in step 104, the travel-related information may be input to the distance model, so as to determine the estimated travel distance between the location to be verified and each reference location according to the output result of the distance model.

For the convenience of understanding, the following takes an online instant distribution scenario as an example, and a detailed description is given to a specific implementation of the above steps.

In an online instant delivery scenario, a user orders (e.g., orders using a smartphone app or a website) a merchant's goods online via a digital platform (i.e., an online delivery service provider), the platform generates a delivery order for the goods, and specifies a rider to take the delivery order, go to a merchant's physical store by the rider to pick up a meal, and then deliver the goods to the user. If the user receives the item later than the time limit (e.g., 30 minutes), the delivery order is determined to be out of time and the platform needs to reimburse the user. Meanwhile, the platform will track the real-time status of the order (i.e. order status data) and record it for accounting and presentation to the user for better user experience. The order status data records four main events from order placement to delivery by the rider: (1) accepting the order, (2) arriving at the merchant, (3) leaving the merchant (taking the goods), (4) delivering to the user. Since the rider receives a large number of orders every day for distribution, especially during peak hours, the rider usually takes the corresponding goods from different merchants and delivers them to the user one by one. Therefore, for two adjacent orders taken by the rider, the travel time of the rider between the store shops corresponding to the two orders can be obtained according to the departure time of one order and the arrival time of the next order. In other words, in the case where the above-mentioned place to be checked and the reference point are physical stores, the travel time information between any two physical stores (the first physical store and the second physical store) can be obtained by: the method comprises the steps of obtaining the store leaving time information of any delivery order accepted by a delivery party and the store arriving time information of the next delivery order, and determining the travel time information between a first entity store corresponding to the delivery order and a second entity store corresponding to the next delivery order (of the delivery order) according to the store leaving time information and the store arriving time information.

Then, for a large merchant pair, the travel time between a pair of consecutively visited merchant stores is related to the travel distance between the merchant stores. Based on this correlation, the travel time may be used to estimate a travel distance between any two merchants. The trip distance obtained from the trip time is more accurate than the trip distance obtained by registering the location. Based on this, travel time between physical stores of any two merchants can be inferred by arrival time and departure time obtained from order status data. Then, the travel distance between two merchants is deduced by using the travel time between the two merchants, so that whether the registered position of one merchant is suspicious is cross-verified by using the travel distance from the merchant to the nearby merchants.

The manner of acquiring order status data is described below.

Relying on manual collection strategies of a rider smartphone. The rider can manually report the order status during the delivery process through a delivery client program installed on the cell phone. Such as time to arrive at, leave, deliver to the user, etc.

An automatic collection strategy based on a physical Beacon check-in system. To automatically obtain the time for the rider to arrive at and depart from the merchant store, a Beacon device may be deployed within the merchant store. Based on the fact that the shop is equipped with the Beacon equipment, the ID of the Beacon equipment corresponds to the ID of the merchant one by one. Beacon equipment lasts outside broadcast bluetooth signal, when the smart mobile phone of rider is close to Beacon equipment, can detect its bluetooth signal who sends. Then, the smartphone of the rider uploads (under the condition of obtaining the authorization of the rider) the timestamp of the Beacon device and the Beacon device ID to the server, which indicates that the rider is around the store corresponding to the Beacon device at the moment, so that the time of the rider arriving at the store and leaving the store can be automatically acquired.

An automatic collection strategy based on a virtual Beacon check-in system. A virtual Beacon system may be deployed on a delivery client on the merchant side, which is part of an application that the merchant uses to manage orders received on the platform. After obtaining the agreement of the merchant, the virtual Beacon Bluetooth broadcast module is added into the application program of the merchant. When the rider approaches the merchant store, the smart phone of the rider receives the virtual Beacon message generated by the smart phone of the merchant. The collection mechanism is the same as the physical Beacon device described above. The virtual Beacon message does not contain continuous GPS tracking to protect the privacy of the merchant, but relies on the dynamic ID mapping with the merchant ID on the server to correspond to the merchant for the virtual Beacon's store-to-store detection.

Based on the above collection strategy, the following data in the time dimension can be obtained: t1, manual capture timestamp (Man) from rider input; t2, timestamp (Phy) from physical Beacon. From the records from the physical Beacon device, an accurate time stamp of the rider as he passes by the merchant can be automatically obtained. The first time stamp of the same merchant monitored by the rider is regarded as that the rider arrives at the merchant, and the last time stamp is regarded as that the rider leaves the merchant; t3, time stamp (Vir) from virtual Beacon. The way of obtaining the timestamp from the virtual Beacon device is the same as that of the physical Beacon device.

Since the travel time information is related to the location corresponding to the travel, it is necessary to collect data in a spatial dimension, that is, position information of each merchant store, in addition to the time dimension. Based on the above collection strategy, the following data in the spatial dimension can be obtained: l1, GPS position of rider (GPS). When the rider reports arrival or departure on his application (i.e., manually collects timestamps) or his application detects a Beacon signal (i.e., an automatic collection timestamp based on a physical or virtual Beacon system), the application will acquire his GPS record, at which time the rider's location can be considered the location of the merchant store; l2, registered location (Reg) of merchant, which location can be obtained by matching merchant ID uploaded by the rider (from manually reported or detected Beacon device) with merchant ID in database, each merchant ID being associated with a unique location registered by the merchant; l3, the actual location of the merchant, can be manually taken to the field for research, such as visiting some merchant stores with abnormal orders on a regular basis.

Location errors can be identified and the actual location of the merchant outlet inferred based on the data in the temporal and spatial dimensions described above. As shown in FIG. 3, different combinations of spatiotemporal measurements may be made on the collected temporal and spatial data, and for each combination a distance model may be trained that is used to estimate travel distances between nearby merchants. Then, the merchants are positioned according to the travel distance output by the model and compared with the registered positions of the merchants to find out the merchant stores with wrong positions.

Reg + Man, using manually entered order status data (T1) and merchant registration location (L2). Reg + Vir, records obtained using a virtual Beacon device (T3) and merchant registration location (L2). Reg + Phy, records obtained using a physical Beacon device (T2) and merchant registration location (L2). GPS + Phy, records acquired using a physical Beacon device (T2) and rider GPS position (L1). GPS + Man, using the rider GPS position (L1) and manually entered order status data (T1). GPS + Vir, records obtained using rider GPS position (L1) and virtual Beacon device (T3).

In view of the inherent homogeneity of the spatio-temporal data, the distance estimation model can take any one of the five measurements as input, thereby outputting the travel distance. The travel distance is related to different types of features, such as temporal features, spatial features, and personalized features. The time characteristics reflect primarily time-related information including travel time, rush hour, latest arrival time, etc. The spatial characteristics reflect regional information including location, order quantity per unit area, delivery difficulty, and the like. The personalized features reflect the behavior and preferences of the rider, such as carrying orders, historical overtime orders, historical trip distances, etc.

For example, the time information may include at least one of: travel time, peak time period, longest delivery time, shortest estimated travel time, historical average delivery time, historical delivery time of a team, committed delivery time, time slices, historical meal delivery time of a merchant, historical meal fetching time of the merchant, and historical meal delivery time of the merchant. The spatial information may include at least one of: distribution difficulty, likely delay scale, merchant popularity, number of completed orders in the last 30 minutes, mall ID. The personalized information may include at least one of: the order taking amount, the rider grade, the overtime order amount, the rider order taking duration, the average historical travel distance, the rider heat and the historical delayed meal delivery rate of the merchant.

In summary, in the case that the location to be verified and the reference point are physical stores, the travel related information between the location to be verified and any physical store may further include at least one of the following: the method comprises the steps of checking the number of delivery orders contained in the area where a place to be checked belongs to, the number of delivery orders contained in the area where any physical store belongs to, road condition information between the place to be checked and any physical store, current delivery state information of a delivery party going to another place from one place of the place to be checked and any physical store, and delivery preference information of the delivery party.

Based on the selected features, because the GBDT can learn the nonlinear relationship between the features and the labels, a distance model can be obtained based on GBDT (Gradient Boosting Decision Tree) training. Also, as a decision tree model, the GBDT may interpret the importance of features, revealing the rider's movement behavior pattern. The labels, outputs and loss functions of the distance model are presented below.

Labeling: distance d of map journey_map(p, q), the travel distance between merchant p and merchant q calculated in the mapping application may be referred to as "map travel distance" d_map(p, q) and training it as a label to the supervised GBDT model. Consider that different travel modes (e.g. walking or electric bicycle) correspond to notWith the same travel distance, two models can be trained respectively: one is for shorter distance walking models (e.g. walking model)<100m) and the other is an electric bicycle model for a longer distance (e.g., 100 m). In this regard, dbike may be set as a threshold for long distances to determine which model should be selected for the sample data. Of course, in order to ensure that the prediction of the distance model is accurate enough, the GPS data of the rider can be collected in the distribution process under the condition of obtaining the authorization of the rider, so that the actual travel distance can be obtained as a label by using the GPS data.

And (3) outputting: model estimation distance d_model(p, q), recording the distance between the commercial tenant p and the commercial tenant q predicted by the distance model as the model estimated distance d_model(p,q)。

Loss function: MAE (Mean Absolute Error) was used as a loss function for training the model.

And 106, positioning the place to be verified according to the declaration position of each reference place and the estimated travel distance to obtain the estimation position of the place to be verified, matching the estimation position with the declaration position of the place to be verified, and judging whether the declaration position of the place to be verified passes verification according to the matching result.

When the to-be-verified location is located according to the declaration position of each reference location and the estimated travel distance, an error condition can be set, and then the target location is searched, so that the error between the standard travel distance between the position of the target location and the declaration position of each reference location and the estimated travel distance meets the preset error condition, and the position of the target location is used as the estimated position. For example, the error condition may be set to have an error within an error threshold, or a target location with the smallest error may be selected as the estimated location. Of course, the specific form of the error condition can be flexibly set according to the actual situation, and the description does not limit this.

Specifically, the target location may be found by:

Similarly, the standard trip distance may include a trip distance between the position of the target location calculated using the map data and the declared position of each reference location.

For ease of understanding, the following description will be continued with an example that is carried out on the above distance model.

Based on the trained model estimated distance, each merchant store can be located and checked for significant deviation from the registered location. For commercial tenant q^*Referred to as a "target merchant" (registered location may be fraudulent), given a set of merchants P in its vicinity ═ P_iAnd i is 1 … N, which is called a reference merchant. d_model(p_i,q^*) Is obtained from the user q by a distance model^*To merchant p_iThe model of (2) estimates the distance (i.e., the output of the distance model). Then the goal is to target locations P and d according to these reference merchants_model(p_i,q^*) Estimating a target merchant location

Then will calculate

And q is^*A comparison is made to determine if there is a significant difference between the two based on the comparison (e.g.,>200 m). The above process is performed for each merchant in sequence, that is, each merchant is traversed, and each merchant is a target merchant once and is also a reference merchant of a nearby merchant. In particular, let G ═ (V, E) be the road network, where V is the set of road intersections and E ═ E { (E)_v1，v2:v₁,v₂E.v is the set of road segments. As shown in FIG. 4, given two merchants p₁And p₂，d_map(p₁,p₂) Is the map trip distance between the two. E.g. d_map(p₁,p₂) Is equal to d_map(p₁,v₁)+|e_v1，v2|+d_map(v₂,p₂) Wherein | e_v1，v2I is the length of the road section, d_map(p₁,v₁) i is the side length | e_p1，v1L. Note that, since the rider rarely detours, the present specification specifies the shortest route when calculating the map trip distance. Thus, d_map(p₁,p₂) Corresponding to the route (p)₁,v₁,v₂,p₂) Rather than (p)₁,v₄,v₃,p₂)。

For the above defined objectives, the problem can be described as: given a set of reference merchants (P ═ { P ═ P)_iI 1 … N) and the model estimate distance d)_model(p_i,q^*) I 1 … N, the merchant location problem is found

So that the model estimates the distance d_model(p_i,q^*) Distance to map journey

Has the smallest error for all p_iE.g. P, i

This optimal estimated position

Should register location q with the merchant^*And (4) approaching. If not, it is detected as a suspicious merchant and manually verified. The above problem is that the travel distance between merchants is limited based on the travel distance of the merchant constrained on the road network, and no connecting edge exists between merchants. Considering that the problem is based on travel distance between merchants on the road network, this specification refers to it as "graph-based multilateration".

In order to solve the graph-based multilateration problem, the present specification designs a road segment search algorithm, whose basic idea is to scan all road segments and search for an optimal location for each road segment. The solution in the algorithm is globally optimal over polynomial time, which allows it to be extended to periodic large-scale detection, e.g., once per day or per week, to reduce or minimize delays in fraud detection.

Road section searching algorithm:

input: road network G ═ (V, E); reference merchant P ═ { P_iI ═ 1 … N }; model estimation distance d_model(p_i,q^*)

Output: estimating a location

min_g←∞

for

do

if g(q)<min_g then

min_g＝g(q),

Where g (q) is a convex function with respect to q (d (u, p)) to be solved directly by Newton's method. Each edge e is iterated, so the computational complexity of the algorithm is

I.e. a polynomial.

It should be noted that merchants near the target merchant can mutually verify the correctness of the location, thereby assisting in positioning. And the pairs of merchants far away have fewer rider tracks and therefore provide less information. More importantly, long range runs may introduce more varying noise to the data. Thus, the distance model and the positioning algorithm can be run in a small range.

The above-mentioned small-range setting can be achieved by clustering the merchants into small subsets, with distance estimation and positioning being performed within each cluster, and all merchants can be clustered based on: urban and suburban areas are divided; the intra-cluster and inter-cluster distances are significantly different, for example, the intra-cluster distance is 2KM, and the inter-cluster distance is at least 5 KM; the radius of coverage of a single cluster is no greater than 2 KM.

The business is clustered by using a hierarchical clustering algorithm BIRCH, wherein the BIRCH is a widely adopted clustering algorithm, clustering is carried out from bottom to top, clustering is carried out from a single node, a point with the minimum distance between clusters is iteratively searched, and if the distance is close enough, the point is combined. When all iterations are over, clusters of fewer than 4 merchants are merged into nearby clusters, since at least three points are needed to locate the fourth point. The description defines the center point P-of the merchant cluster P as its average location. And the radius r (P) of the cluster P is the farthest distance traveled by the merchant to the center point. Given two merchant clusters P_iAnd P_jDistance between clusters D_out(P_i,P_j) Is defined as P_iAnd P_jThe minimum trip distance between merchants.

And (3) clustering of merchants:

input: commercial tenant set N ═ { p_i}

Output: set of merchant clusters

where P_i←{p_i}

while

Find P_i,

With minimum in-class distance

if D_out(P_i,P_j)>αthen

Break；

Will P_i,P_jAre merged into a cluster P_*

If | P_i|<4, merging P_iTo the nearest cluster.

As can be seen from the above embodiments, the position error of the outdoor commercial tenant is automatically sensed and corrected based on the order state data of the online instant distribution platform, so that additional labor cost is not required, position correction can be completed under the condition that a rider feels no, and low-cost large-scale deployment is facilitated.

Corresponding to the method embodiment, the specification also provides a corresponding device embodiment.

FIG. 5 is a schematic block diagram of an apparatus provided in an exemplary embodiment. Referring to fig. 5, at the hardware level, the apparatus includes a processor 502, an internal bus 504, a network interface 506, a memory 508 and a non-volatile memory 510, but may also include hardware required for other services. One or more embodiments of the present description may be implemented in software, such as by processor 502 reading corresponding computer programs from non-volatile storage 510 into memory 508 and then running. Of course, besides software implementation, the one or more embodiments in this specification do not exclude other implementations, such as logic devices or combinations of software and hardware, and so on, that is, the execution subject of the following processing flow is not limited to each logic unit, and may also be hardware or logic devices.

Referring to fig. 6, the position verification apparatus may be applied to the device shown in fig. 5 to implement the technical solution of the present specification. Wherein, the position verifying unit may include:

a determining unit 61, configured to determine a location to be verified and a plurality of reference locations corresponding to the location to be verified, and obtain declaration positions of the location to be verified and the reference locations;

the pre-estimating unit 62 is used for determining the estimated travel distance between the to-be-verified place and each reference place according to travel related information between the to-be-verified place and each reference place, wherein the travel related information comprises travel time information between the to-be-verified place and each reference place;

the verification unit 63 is configured to locate the place to be verified according to the declaration position of each reference place and the estimated travel distance to obtain an estimated position of the place to be verified, match the estimated position with the declaration position of the place to be verified, and determine whether the declaration position of the place to be verified passes verification according to a matching result.

Optionally, the determining unit 61 is specifically configured to:

Optionally, the estimating unit 62 is specifically configured to:

Optionally, the estimating unit 62 is further configured to:

Optionally, the verification unit 63 is specifically configured to:

taking the position of the target location as the estimated position.

Optionally, the target location is searched in the following manner:

The systems, devices, modules or units illustrated in the above embodiments may be implemented by a computer chip or an entity, or by a product with certain functions. A typical implementation device is a computer, which may take the form of a personal computer, laptop computer, cellular telephone, camera phone, smart phone, personal digital assistant, media player, navigation device, email messaging device, game console, tablet computer, wearable device, or a combination of any of these devices.

In a typical configuration, a computer includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.

The memory may include forms of volatile memory in a computer readable medium, Random Access Memory (RAM) and/or non-volatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). Memory is an example of a computer-readable medium.

Computer-readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), Read Only Memory (ROM), Electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), Digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic disk storage, quantum memory, graphene-based storage media or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device. As defined herein, a computer readable medium does not include a transitory computer readable medium such as a modulated data signal and a carrier wave.

It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.

The foregoing description has been directed to specific embodiments of this disclosure. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims may be performed in a different order than in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing may also be possible or may be advantageous.

The terminology used in the description of the one or more embodiments is for the purpose of describing the particular embodiments only and is not intended to be limiting of the description of the one or more embodiments. As used in one or more embodiments of the present specification and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It should also be understood that the term "and/or" as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items.

It should be understood that although the terms first, second, third, etc. may be used in one or more embodiments of the present description to describe various information, such information should not be limited to these terms. These terms are only used to distinguish one type of information from another. For example, first information may also be referred to as second information, and similarly, second information may also be referred to as first information, without departing from the scope of one or more embodiments herein. The word "if" as used herein may be interpreted as "at … …" or "when … …" or "in response to a determination", depending on the context.

The above description is only for the purpose of illustrating the preferred embodiments of the one or more embodiments of the present disclosure, and is not intended to limit the scope of the one or more embodiments of the present disclosure, and any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the one or more embodiments of the present disclosure should be included in the scope of the one or more embodiments of the present disclosure.

Claims

1. A method for location verification, comprising:

2. The method of claim 1, wherein determining the location to be verified and a number of reference locations corresponding to the location to be verified comprises:

3. The method according to claim 1, wherein the determining the estimated travel distance between the location to be verified and each reference location according to the travel related information between the location to be verified and each reference location respectively comprises:

4. The method according to claim 1, wherein in case the location to be verified and the reference point are physical stores, the method further comprises:

5. The method of claim 1, wherein the positioning the location to be verified according to the declared position and the estimated travel distance of each reference location to obtain the estimated position of the location to be verified comprises:

taking the position of the target location as the estimated position.

6. The method of claim 5, wherein the target location is located by:

7. The method of claim 5, wherein the standard trip distance comprises a trip distance between the location of the target location and a declared location of each reference location calculated using map data.

8. A position verification apparatus, comprising:

9. An electronic device, comprising:

a processor;

a memory for storing processor-executable instructions;

wherein the processor implements the method of any one of claims 1-7 by executing the executable instructions.

10. A computer-readable storage medium having stored thereon computer instructions, which when executed by a processor, perform the steps of the method according to any one of claims 1-7.