Address identification method and device, equipment and storage medium

Abstract

The embodiment of the invention discloses an address identification method and device, equipment and a storage medium. The method comprises the following steps: acquiring a historical delivery task of an abnormal object; based on the historical delivery task, determining a target historical delivery address meeting a preset stable delivery condition from the historical delivery addresses of the abnormal objects; based on a target historical delivery task corresponding to the target historical delivery address, determining a delivery address association graph taking each target historical delivery address as a center; determining a target abnormal degree corresponding to each target historical shipping address based on the shipping address association graph; and determining a target shipping address of a normal object served by the abnormal object based on the target abnormal degree corresponding to the target historical shipping address. Through the technical scheme of the embodiment of the invention, automatic identification of the shipping address can be realized, manual participation is not needed, the address identification efficiency is greatly improved, and the objectivity and stability of an identification result are ensured.

Description

Technical Field

[0001] Embodiments of the present invention relate to computer technology, and in particular, to an address recognition method, apparatus, device, and storage medium. Background Art

[0002] In the logistics industry, there is an abnormal object, such as a scalper merchant. This abnormal object cooperates with the logistics service provider, uses the service to pick up the items that normal objects need to ship, and makes a profit from the price difference. This abnormal object will damage the interests of the logistics service provider, so when such an abnormal object is found, the cooperation will be stopped in time. However, the normal objects served by this abnormal object are potential customers of the logistics service provider. Therefore, it is necessary to identify the normal objects served by this abnormal object in order to convert the normal objects into direct customers of the logistics service provider, and additional income can be increased while minimizing losses.

[0003] Currently, it is usually determined whether the shipping address is the address where the normal object served by the abnormal object is located by means of manual on-site visits to the shipping address of the abnormal object.

[0004] However, in the process of implementing the present invention, the inventor found that there are at least the following problems in the prior art:

[0005] The existing on-site visit and recognition method is time-consuming and laborious, greatly reducing the address recognition efficiency, and relying on subjective human judgment, which cannot guarantee the objectivity and stability of the recognition result. Summary of the Invention

[0006] Embodiments of the present invention provide an address recognition method, apparatus, device, and storage medium to realize automatic recognition of shipping addresses without manual participation, greatly improving the address recognition efficiency and ensuring the objectivity and stability of the recognition result.

[0007] In a first aspect, an embodiment of the present invention provides an address recognition method, including:

[0008] Obtain the historical shipping tasks of the abnormal object;

[0009] Based on the historical shipping tasks, determine target historical shipping addresses that meet the preset stable shipping conditions from the historical shipping addresses of the abnormal object;

[0010] Based on the target historical shipping tasks corresponding to the target historical shipping addresses, determine a shipping address association graph centered on each of the target historical shipping addresses;

[0011] Based on the shipping address association graph, determine the target abnormal degree corresponding to each of the target historical shipping addresses;

[0012] Determine the target delivery address of the normal object served by the abnormal object based on the target abnormality degree corresponding to the target historical delivery address.

[0013] In a second aspect, an embodiment of the present invention further provides an address recognition device, including:

[0014] A historical delivery task acquisition module, configured to acquire the historical delivery tasks of the abnormal object;

[0015] A target historical delivery address determination module, configured to determine, based on the historical delivery tasks, a target historical delivery address that meets the preset stable delivery condition from the historical delivery addresses of the abnormal object;

[0016] A delivery address association graph determination module, configured to determine a delivery address association graph centered on each target historical delivery address based on the target historical delivery task corresponding to the target historical delivery address;

[0017] A target abnormality degree determination module, configured to determine the target abnormality degree corresponding to each target historical delivery address based on the delivery address association graph;

[0018] A target delivery address determination module, configured to determine the target delivery address of the normal object served by the abnormal object based on the target abnormality degree corresponding to the target historical delivery address.

[0019] In a third aspect, an embodiment of the present invention further provides an electronic device, where the electronic device includes:

[0020] One or more processors;

[0021] A memory, configured to store one or more programs;

[0022] When the one or more programs are executed by the one or more processors, the one or more processors implement the address recognition method provided in any embodiment of the present invention.

[0023] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, it implements the address recognition method provided in any embodiment of the present invention.

[0024] One embodiment of the above invention has the following advantages or beneficial effects:

[0025] Based on the historical shipping tasks of the exception object, determine the target historical shipping address that meets the preset stable shipping conditions from the historical shipping addresses of the exception object, and based on the target historical shipping tasks corresponding to the target historical shipping address, determine the shipping address association graph centered on each target historical shipping address. Based on the shipping address association graph, determine the target exception degree corresponding to each target historical shipping address, and based on the target exception degree corresponding to the target historical shipping address, determine the target shipping address of the normal object served by the exception object, so as to realize the automatic identification of the shipping address of the normal object served by the exception object, without manual participation, greatly improving the address identification efficiency and ensuring the objectivity and stability of the identification result. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0027] Figure 1 is a flowchart of an address recognition method provided by an embodiment of the present invention;

[0028] Figure 2 is a flowchart of another address recognition method provided by an embodiment of the present invention;

[0029] Figure 3 is an example of a shipping address association graph involved in an embodiment of the present invention;

[0030] Figure 4 is a flowchart of yet another address recognition method provided by an embodiment of the present invention;

[0031] Figure 5 is a schematic structural diagram of an address recognition device provided by an embodiment of the present invention;

[0032] Figure 6 is a schematic structural diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0033] The following will further elaborate on the present invention in conjunction with the drawings and embodiments. It can be understood that the specific embodiments described herein are only used to explain the present invention, rather than limiting the present invention. Additionally, it should be noted that for the sake of convenience of description, only the parts related to the present invention are shown in the drawings, rather than all the structures.

[0034] It should be noted that in the technical solution of the present disclosure, in terms of the collection, gathering, updating, analysis, processing, use, transmission, storage, etc. of the user's personal information, it complies with the provisions of relevant laws and regulations, is used for legal purposes, and does not violate public order and good customs. Necessary measures are taken for the user's personal information to prevent illegal access to the user's personal information data, and to safeguard the security of the user's personal information, network security, and national security.

[0035] Figure 1 FIG. is a flowchart of an address recognition method provided by an embodiment of the present invention. This embodiment is applicable to the situation of identifying the shipping address of a normal object served by an abnormal object. This method can be executed by an address recognition device, which can be implemented in software and / or hardware and integrated into an electronic device. As Figure 1 shown, the method specifically includes the following steps:

[0036] S110. Obtain the historical shipping tasks of the abnormal object.

[0037] Among them, the abnormal object can refer to an intermediate party between the service provider and the service demander. The abnormal object is not the real logistics service demander, that is, it is not the object that truly has the shipping demand. Relative to the abnormal object, the normal object refers to the service demander who truly has the shipping demand. The normal object is the upstream customer of the abnormal object. The abnormal object can serve multiple normal objects, and the normal objects ship through the abnormal object. For example, the abnormal object can refer to a scalper merchant in the logistics industry. The scalper merchant cooperates with the logistics service provider and collects the items that the served normal objects need to ship, and earns the price difference. When the abnormal object ships, it can directly ship with the location of the normal object as the shipping address, or it can also concentrate the goods of the normal object to certain places, that is, the goods collection place for shipping. It should be noted that the normal objects served by the abnormal object can include upstream customers with sporadic shipping demands and upstream customers with stable shipping demands. Among them, the sporadic shipping demand can refer to having a small amount of shipping demand within a period of time. For example, the sporadic shipping demand can refer to a demand with a shipping frequency less than the pre-set shipping frequency or can also refer to a demand with relatively discrete shipping times within a period of time. The stable shipping demand can refer to having a large amount of continuous shipping within a period of time. For example, the stable shipping demand can refer to a demand with a shipping frequency greater than the preset shipping frequency or can also refer to a demand with relatively fixed shipping times within a period of time, such as the demand for shipping every day. The upstream customers with stable shipping demands are potential customers of the logistics service provider. Therefore, it is necessary to identify the shipping addresses of the normal objects with stable shipping demands through the abnormal object, so as to convert these normal objects into the own customers of the logistics service provider, thereby increasing revenue while stopping losses.

[0038] Among them, the historical delivery tasks can refer to the delivery orders generated when abnormal objects are delivered within a historical time period. The number of historical delivery tasks is multiple, and each historical delivery task can include, but is not limited to: delivery attribute information, delivery item information, and delivery task generation time. Among them, the delivery attribute information can include, but is not limited to: delivery address, delivery party name, delivery mobile phone number, and delivery application account. The delivery item information can include, but is not limited to: delivery item category, delivery item weight, and delivery item volume. The delivery task generation time can refer to the delivery order placement time. The delivery application account can refer to the delivery order placement account.

[0039] Specifically, after the service provider discovers an abnormal object, such as discovering a scalper merchant, the delivery address of the normal object served by the abnormal object can be automatically identified by executing steps S110 - S150. For example, the identification device can obtain all the historical delivery tasks issued by the delivery object from the delivery server within the most recent historical time period.

[0040] S120. Based on the historical delivery tasks, determine the target historical delivery address that meets the preset stable delivery conditions from the historical delivery addresses of the abnormal object.

[0041] Among them, the historical delivery address can refer to the delivery address in the historical delivery task. Stable delivery can refer to the delivery frequency and / or delivery time being relatively fixed, that is, the fluctuation of the delivery frequency or the delivery time is within a preset range. For example, stable delivery can refer to the delivery frequency being greater than the preset delivery frequency, or the delivery time being fixed within a period of time, such as delivering every day. The preset stable delivery conditions can refer to the conditions that need to be met for preset stable delivery. For example, the preset stable delivery conditions can refer to the delivery frequency being greater than the preset delivery frequency, or it can also refer to the delivery time being fixed within a period of time. The target historical delivery address can refer to the historical delivery address with stable delivery requirements. The target historical delivery address can include the delivery addresses of normal objects with stable delivery requirements, and can also include the goods collection addresses of abnormal objects.

[0042] Specifically, feature extraction, cleaning, and processing can be performed on the task information in each historical shipping task to obtain the shipping feature information corresponding to each historical shipping address of the abnormal object. And through feature comparison or clustering, the target historical shipping addresses that meet the preset stable shipping conditions can be determined based on the shipping address information. For example, the target historical shipping addresses with stable shipping requirements can be determined by detecting whether the shipping address information meets the preset stable shipping conditions. For example, the shipping feature information may include, but is not limited to: the number of historical shipping tasks, the historical shipping frequency, the number of shipping mobile phone numbers used, and the number of shipping application accounts. If the historical shipping address in the historical shipping task meets at least one of the conditions that the number of historical shipping tasks is greater than or equal to the preset shipping quantity threshold, the historical shipping frequency is greater than the preset shipping frequency threshold, the number of shipping mobile phone numbers used is less than or equal to the preset mobile phone number threshold, and the number of shipping application accounts is less than or equal to the preset account number threshold, then the historical shipping address is determined as the target historical shipping address with stable shipping requirements. By screening out the target historical shipping addresses that meet the preset stable shipping conditions from all historical shipping tasks, the shipping addresses of normal objects with stable shipping requirements and the collection addresses of abnormal objects can be identified, and the historical shipping addresses of scattered upstream customers are excluded.

[0043] S130. Based on the target historical shipping tasks corresponding to the target historical shipping addresses, determine the shipping address association graph centered on each target historical shipping address.

[0044] Among them, the target historical shipping task may refer to the historical shipping task with the target historical shipping address as the shipping address. Each target historical shipping address corresponds to a shipping address association graph. The shipping address association graph can be used to represent the association relationship between the target historical shipping address in the middle position and other target historical shipping addresses. For example, the shipping address association graph may refer to the Ego-Network. This network node consists of a unique central node Ego and the neighbor nodes Alters of this central node. The edges only include the connections between the central node Ego and the neighbor nodes Alter and the connections between the neighbor nodes Alter.

[0045] Specifically, each target historical shipping address can be abstracted as a node. For each target historical shipping address, based on the target historical shipping task corresponding to each target historical shipping address, it can be determined whether there is an association between the current target historical shipping address and other target historical shipping addresses, and whether there is an association between other target historical shipping addresses. And the nodes corresponding to the target historical shipping addresses with associations are connected by lines, so as to construct a shipping address association graph centered on the current target historical shipping address, and the association size between the target historical shipping addresses with connections, that is, the edge weight, can be determined.

[0046] S140. Determine the target abnormality degree corresponding to each target historical shipping address based on the shipping address association graph.

[0047] Among them, the target abnormality degree can be used to characterize the likelihood that the target historical shipping address is the goods collection address of an abnormal object. Specifically, for the shipping address association graph corresponding to each target historical shipping address, the abnormal nodes in the current shipping address association graph can be detected based on the unsupervised method of graph anomaly point detection, and the target abnormality degree that the current target historical shipping address is the goods collection address can be obtained.

[0048] Exemplarily, S140 may include: determining the abnormality degree under each abnormal structure based on at least one abnormal structure and the current shipping address association graph centered on the current target historical shipping address; determining the target abnormality degree corresponding to the current target historical shipping address based on the abnormality degree under each abnormal structure.

[0049] Among them, the abnormal structure may refer to the abnormal pattern existing in the shipping address association graph. For example, the abnormal structure may include, but is not limited to, at least one of a first abnormal structure in the shape of a star or a cluster, a second abnormal structure with an abnormal total edge weight, and a third abnormal structure with an abnormal edge weight. Among them, the first abnormal structure CliqueStar may include a star-shaped abnormal structure Near-Star and a cluster-shaped abnormal structure Near-Clique identified based on density. The star-shaped abnormal structure means that in the shipping address association graph, the central node is associated with neighbor nodes, but there is almost no association between the neighbor nodes, presenting a star-shaped structure. The cluster-shaped abnormal structure means that in the shipping address association graph, the central node is associated with neighbor nodes, and there is a large amount of association between the neighbor nodes, presenting a cluster-shaped structure. The second abnormal structure HeavyVicinity may be that in the shipping address association graph, when the number of edges is certain, the total edge weight is abnormally large, and the edge weight between the central node and some neighbor nodes is also abnormally large. The third abnormal structure DominantPair may be that in the shipping address association graph, there is an edge with an abnormally large edge weight between the central node and a neighbor node. The first abnormal structure, the second abnormal structure, and the third abnormal structure can all be used to measure the abnormality degree that the target historical shipping address is the goods collection address.

[0050] Specifically, each target historical shipping address can be used as the current target historical shipping address to determine the degree of abnormality. For example, in the current shipping address association graph centered on the current target historical shipping address, the degree of abnormality of the current target historical shipping address as a goods collection address can be determined under each abnormal structure. If there is only one abnormal structure, the degree of abnormality under this abnormal structure can be directly determined as the target degree of abnormality corresponding to the current target historical shipping address. If there are at least two abnormal structures, the degrees of abnormality under each abnormal structure can be weighted and summed, and the obtained result is used as the target degree of abnormality corresponding to the current target historical shipping address, so that the degrees of abnormality under multiple abnormal structures can be integrated to further improve the accuracy of address recognition.

[0051] S150. Determine the target shipping address of the normal object served by the abnormal object based on the target degree of abnormality corresponding to the target historical shipping address.

[0052] Among them, the target shipping address can refer to the shipping address of the normal object with stable shipping demand served by the abnormal object. Specifically, the target historical shipping address with a target degree of abnormality less than or equal to the preset abnormality degree threshold can be used as the target shipping address of the normal object served by the abnormal object, so that the shipping addresses of the upstream customers of scalper merchants can be automatically identified. It should be noted that the target historical shipping address with a target degree of abnormality greater than the preset abnormality degree threshold is the goods collection address of the abnormal object. By using the shipping address association graph, the shipping addresses of the normal objects served by the abnormal object can be automatically identified from all target historical shipping addresses, avoiding the risk of manual intervention and enhancing the objectivity and stability of address recognition.

[0053] It should be noted that by first screening out the target historical shipping addresses with stable shipping demand and then performing abnormality detection based on the shipping address association graph, both the characteristic performance of the shipping addresses of normal objects themselves and the performance of the shipping addresses of normal objects in the network formed with other shipping addresses are considered, thus avoiding the one-sidedness of using a single recognition method and increasing the characteristic dimension and recognition dimension, ensuring the accuracy of address recognition.

[0054] In the technical solution of this embodiment, based on the historical shipping tasks of the abnormal object, the target historical shipping address that meets the preset stable shipping condition is determined from the historical shipping addresses of the abnormal object. Based on the target historical shipping task corresponding to the target historical shipping address, the shipping address association graph centered on each target historical shipping address is determined. Based on the shipping address association graph, the target abnormal degree corresponding to each target historical shipping address is determined. And based on the target abnormal degree corresponding to the target historical shipping address, the target shipping address of the normal object served by the abnormal object is determined, so as to realize the automatic identification of the shipping address of the normal object served by the abnormal object, without manual participation, greatly improving the address identification efficiency and ensuring the objectivity and stability of the identification result.

[0055] On the basis of the above technical solution, S120 may include: performing information processing on the historical shipping tasks to determine the shipping feature information corresponding to each historical shipping address of the abnormal object; based on the shipping feature information, performing grouping and clustering on the historical shipping addresses to obtain each historical shipping address group after clustering; and based on the average shipping feature information corresponding to the historical shipping address group, determining the target historical shipping address that meets the preset stable shipping condition.

[0056] Among them, the historical shipping feature information corresponding to each historical shipping address may include, but is not limited to: the number of historical shipping tasks, the historical shipping frequency, the historical shipping concentration index information, the historical shipping stability index information, and the historical shipping address uniqueness index information. Among them, the number of historical shipping tasks may refer to the total number of historical shipping tasks with this historical shipping address as the shipping address, that is, the total number of waybills. The historical shipping frequency may refer to the shipping frequency in the recent historical time period, such as the number of shipping days. The historical shipping concentration index information may include, but is not limited to: at least one of the number of different shipping mobile phone numbers and the number of shipping application accounts in the historical shipping tasks with this historical shipping address as the shipping address. The historical shipping stability index information may include, but is not limited to: at least one of the standard deviation of the number of historical shipping tasks, the standard deviation of the historical shipping weight, the standard deviation of the historical shipping volume, and the number of historical shipping categories statistically every preset time. The historical shipping address uniqueness index information may include, but is not limited to: at least one of the number of different shipping mobile phone numbers and the number of shipping application accounts in the historical shipping tasks with this historical shipping address as the shipping address statistically every preset time. The historical shipping address group may be a set composed of multiple similar historical shipping addresses.

[0057] Specifically, feature extraction, cleaning, processing, and normalization can be performed on the task information in each historical shipping task to obtain the shipping feature information corresponding to each historical shipping address of the abnormal object. The shipping feature information can be used as the feature value of the corresponding historical shipping address for unsupervised clustering, such as K-means clustering, to divide all historical shipping addresses into multiple groups of historical shipping addresses. For example, the elbow method can be used in advance to determine the number of clusters K. The elbow method determines the sum of squared errors corresponding to each K value, and by observing the K value that causes the sum of squared errors to suddenly decrease as the K value increases, that is the appropriate number of clusters K. After determining the number of clusters K, K central addresses can be selected from all historical shipping addresses, the distance between the central addresses and other addresses can be determined based on the shipping feature information, such as the Euclidean distance, and based on this distance, other addresses can be divided into the group with the shortest distance to the central address, thereby dividing all historical shipping addresses into k groups of historical shipping addresses, and the addresses in each group of historical shipping addresses have similar performance. The shipping feature information corresponding to each historical shipping address in each group of historical shipping addresses can be averaged to obtain the average shipping feature information corresponding to each group of historical shipping addresses. By comparing the average shipping feature information corresponding to each group of historical shipping addresses, the target group of historical shipping addresses with stable shipping requirements can be determined, and the historical shipping addresses in this target group of historical shipping addresses can be determined as the target historical shipping addresses with stable shipping requirements. For example, if there has been continuous shipping historically and the shipping is relatively concentrated on certain shipping mobile phone numbers or shipping application accounts, it indicates that this address has stable shipping requirements. On the contrary, if the address has only had a small number of shipments or a small volume of goods shipped historically, or there are a large number of mobile phone numbers for temporary shipments, it is likely not an address with stable shipping requirements. Through unsupervised clustering, historical addresses with stable shipping requirements can be more accurately identified.

[0058] Based on the above technical solution, S150 may include: performing grouped clustering on the target historical shipping addresses based on the target abnormal degree corresponding to the target historical shipping addresses to obtain a group of abnormal shipping addresses and a group of non-abnormal shipping addresses; determining the target historical shipping addresses in the group of non-abnormal shipping addresses as the target shipping addresses of the normal objects served by the abnormal object.

[0059] Among them, the group of abnormal shipping addresses may refer to the group where the goods collection address of the abnormal object is located. The group of non-abnormal shipping addresses may refer to the group where the shipping addresses of the normal objects with stable shipping requirements served by the abnormal object are located.

[0060] Specifically, the target abnormality degree corresponding to each target historical shipping address can be used as a feature value for binary clustering, and all target historical shipping addresses are divided into an abnormal shipping address group and a non-abnormal shipping address group, so as to distinguish the goods collection address and the shipping addresses of normal objects, and determine the target shipping addresses of normal objects with stable shipping requirements served by abnormal objects from the target historical shipping addresses in the non-abnormal shipping address group. By using an unsupervised clustering method, the shipping addresses of normal objects can be identified more accurately without threshold setting, thereby further improving the accuracy of address recognition.

[0061] Figure 2 FIG. is a flowchart of another address recognition method provided by an embodiment of the present invention. Based on the above embodiments, the specific determination process of the shipping address association graph is described in detail. The explanations of the same or corresponding terms in the above embodiments are not repeated here.

[0062] See Figure 2 , another address recognition method provided by this embodiment specifically includes the following steps:

[0063] S210. Obtain the historical shipping tasks of the abnormal object.

[0064] S220. Based on the historical shipping tasks, determine the target historical shipping addresses that meet the preset stable shipping conditions from the historical shipping addresses of the abnormal object.

[0065] S230. Based on the target historical shipping tasks corresponding to the target historical shipping addresses, detect whether every two target historical shipping addresses have the same shipping attribute information.

[0066] Among them, the shipping attribute information can be pre-set attribute information for detecting whether there is an association between two shipping addresses. For example, the shipping attribute information may include, but is not limited to, the shipping mobile phone number and / or the shipping application account.

[0067] Specifically, based on the target historical shipping tasks, the shipping attribute information corresponding to each target historical shipping address can be determined, such as the shipping mobile phone number and / or the shipping application account used in the shipping tasks with each target historical shipping address as the shipping address. Compare the shipping attribute information corresponding to every two target historical shipping addresses to determine whether there is the same shipping attribute information, such as whether the same shipping mobile phone number and / or shipping application account is used.

[0068] S240. Take each target historical shipping address as a node, connect the associated nodes that have the same shipping attribute information as the current node, and connect other associated nodes that have the same shipping attribute information as the current associated node, to obtain a shipping address association graph centered on the current node, and determine the edge weights in the shipping address association graph.

[0069] Among them, the current node may refer to the target historical shipping address that is currently the central node. A corresponding shipping address association graph is determined by taking each target historical shipping address as the current node. The associated node may refer to the node where the target historical shipping address with the same shipping attribute information as the current node is located. The current associated node may refer to any associated node specified currently. Other associated nodes may refer to the remaining associated nodes except the current associated node. The edge weight may be used to characterize the degree of association between two nodes connected by an edge.

[0070] Specifically, Figure 3 An example of a shipping address association graph is given. As Figure 3 shown, each target historical shipping address is abstracted as a node, and the target historical shipping address is set as A i (i = 1, 2, 3, ……, k). If the current node A i has the same shipping attribute information as other nodes A j , then determine A j as the associated node, and connect A i and A j . Similarly, if the current associated node A j has the same shipping attribute information as other associated nodes A r , then connect A j and A r . Thus, a shipping address association graph Ego - Network centered on A i is constructed. Based on the shipping attribute information corresponding to the two nodes with a connection, determine the edge weight corresponding to each edge in the shipping address association graph.

[0071] Exemplarily, "determine the edge weights in the shipping address association graph" in S240 may include: obtaining the first historical shipping task corresponding to the first node of the current edge in the shipping address association graph and the second historical shipping task corresponding to the second node; determining the number of target first historical shipping tasks that have the same shipping attribute information as at least one second historical shipping task, and determining the number of target second historical shipping tasks that have the same shipping attribute information as at least one first historical shipping task; based on the number of target first historical shipping tasks, the number of target second historical shipping tasks, the total number of first historical shipping tasks, and the total number of second historical shipping tasks, determine the edge weight corresponding to the current edge.

[0072] Among them, the delivery address in the first historical delivery task is the target historical delivery address corresponding to the first node. The delivery address in the second historical delivery task is the target historical delivery address corresponding to the second node. The current edge can be any edge of the delivery address association graph, and the corresponding edge weight is determined by taking each edge in the delivery address association graph as the current edge. The first node and the second node are respectively the two nodes connected by the current edge.

[0073] Specifically, the first historical delivery task with the target historical delivery address A corresponding to the first node as the delivery address can be obtained i and the second historical delivery task with the target historical delivery address A corresponding to the second node as the delivery address. In the first historical delivery task, the target first historical delivery task that uses the same delivery mobile phone number and / or delivery application account as at least one second historical delivery task is determined, and the number N of the target first historical delivery tasks is obtained j . Similarly, in the second historical delivery task, the target second historical delivery task that uses the same delivery mobile phone number and / or delivery application account as at least one first historical delivery task is determined, and the number N of the target second historical delivery tasks is obtained ij . Based on the number N of the target first historical delivery tasks ji , the number N of the target second historical delivery tasks ij , the total number N of the first historical delivery tasks ji and the total number N of the second historical delivery tasks i , the association degree between the first node and the second node, that is, the edge weight, is determined. j

[0074] Exemplarily, based on the number of the target first historical delivery tasks, the number of the target second historical delivery tasks, the total number of the first historical delivery tasks, and the total number of the second historical delivery tasks, determining the edge weight corresponding to the current edge may include: adding the number N of the target first historical delivery tasks ij and the number N of the target second historical delivery tasks ji to obtain the number of associated tasks corresponding to the current edge; adding the total number N of the first historical delivery tasks i and the total number N of the second historical delivery tasks j to obtain the total number of delivery tasks corresponding to the current edge; taking the ratio between the number N of the associated tasks corresponding to the current edge ij +N ji and the corresponding total number N of the delivery tasks i +N j as the edge weight corresponding to the current edge.

[0075] S250. Determine the target abnormal degree corresponding to each target historical shipping address based on the shipping address association graph.

[0076] S260. Determine the target shipping address of the normal object served by the abnormal object based on the target abnormal degree corresponding to the target historical shipping address.

[0077] In the technical solution of this embodiment, by detecting whether each two target historical shipping addresses have the same shipping attribute information, connecting the associated nodes with the same shipping attribute information as the current node, and connecting other associated nodes with the same shipping attribute information as the current associated node, the shipping address association graph centered on the current node can be accurately obtained, further improving the accuracy of address recognition.

[0078] Figure 4 It is a flowchart of another address recognition method provided by an embodiment of the present invention. Based on the above embodiments, the specific process of determining the abnormal degree of the shipping address based on three abnormal structures, namely, the first abnormal structure in the shape of a star or a cluster, the second abnormal structure with an abnormal total edge weight, and the third abnormal structure with an abnormal edge weight, is described in detail. The explanations of the same or corresponding terms in the above embodiments are not repeated here.

[0079] See Figure 4 , another address recognition method provided by this embodiment specifically includes the following steps:

[0080] S410. Obtain the historical shipping tasks of the abnormal object.

[0081] S420. Based on the historical shipping tasks, determine the target historical shipping addresses that meet the preset stable shipping conditions from the historical shipping addresses of the abnormal object.

[0082] S430. Based on the target historical shipping tasks corresponding to the target historical shipping addresses, determine the shipping address association graph centered on each target historical shipping address.

[0083] S440. Based on the current number of nodes and the current number of edges in the current shipping address association graph corresponding to the current target historical shipping address, determine the first abnormal degree under the first abnormal structure.

[0084] Among them, the current target historical shipping address can be any one of the target historical shipping addresses, and the corresponding target abnormal degree is determined by taking each target historical shipping address as the current target historical shipping address. The current number of nodes can refer to the total number of nodes in the current shipping address association graph. The current number of edges can refer to the total number of edges in the current shipping address association graph.

[0085] It should be noted that the star-shaped abnormal structure in the first abnormal structure tends to conform to the performance of the goods collection address of the abnormal object, that is, the goods collection address of the abnormal object is associated with other delivery addresses through the delivery application account and the delivery mobile phone number, but there is rarely any association between the other delivery addresses so associated. The central node of this structure is very likely to correspond to the goods collection address of the abnormal object. The cluster-shaped abnormal structure in the first abnormal structure tends to conform to the performance of the group of goods collection addresses of the abnormal object.

[0086] Specifically, the first abnormal structure can measure the degree of abnormality by using the number of nodes and the number of edges, so as to obtain the first degree of abnormality of the current delivery address association graph under the first abnormal structure.

[0087] Exemplarily, S440 may include: determining the fitted number of edges corresponding to the current number of nodes based on the current number of nodes corresponding to the current delivery address association graph centered on the current target historical delivery address and the variation relationship between the number of nodes and the number of edges obtained by pre-fitting; determining the first deviation degree under the first abnormal structure based on the current number of edges and the fitted number of edges corresponding to the current delivery address association graph; determining the first local abnormality degree under the first abnormal structure based on the preset density identification method, the current number of nodes and the current number of edges corresponding to the current delivery address association graph; and determining the first degree of abnormality under the first abnormal structure based on the first deviation degree and the first local abnormality degree.

[0088] Among them, the first deviation degree may refer to the deviation degree between the actual number of edges and the fitted number of edges, that is, the out-line degree. The first local abnormality degree may be the abnormality degree identified based on density, that is, the LOF (Local Outlier Factor) degree, so as to identify the abnormal points that are very close to the fitted value. The preset density identification method may be a pre-configured existing method for identifying the local abnormality degree based on density.

[0089] Specifically, the number of edges and the number of nodes in the delivery address association graph follow a power-law distribution, that is Taking the natural logarithm of the number of edges and the number of nodes respectively, and then fitting by linear regression to obtain the slope and intercept, that is, lnNE = β 1 *lnNA + β 2 , determining the slope β 1 and the intercept β 2 through historical data, and then the variation relationship between the fitted number of nodes and the number of edges can be obtained. Based on this variation relationship and the current number of nodes NA i corresponding to the current delivery address association graph U i , the fitted number of edges can be determined Based on the current number of edges NE i corresponding to the current delivery address association graph and the fitted number of edges Determine the current number of edges NE i and the number of fitted edges Find the maximum and minimum values among them, as well as the difference between the current number of edges and the number of fitted edges, and use log to amplify this difference, then multiply the ratio between the maximum and minimum values by the amplified result to obtain the first deviation degree outline1(i) under the first abnormal structure.

[0090] Based on the current number of nodes and the current number of edges corresponding to the current shipping address association graph, density identification can be performed to determine the first local abnormal degree lof1(i) under the first abnormal structure. The first deviation degree and the first local abnormal degree can be normalized and summed, so as to comprehensively measure the current shipping address association graph U i The first abnormal degree under the first abnormal structure, that is, the first abnormal degree out-score1(i) corresponding to the current target historical shipping address.

[0091] S450. Based on the current number of nodes and the current total edge weight corresponding to the current shipping address association graph, determine the second abnormal degree under the second abnormal structure.

[0092] Among them, the current total edge weight may refer to the sum of the edge weights in the current shipping address association graph. It should be noted that the second abnormal structure also conforms to the performance of the collection address of abnormal nodes, that is, the number of the same shipping application account and shipping mobile phone number used between some shipping addresses is very large. The central node of this structure is very likely to correspond to the collection address of the abnormal object.

[0093] Specifically, the second abnormal structure can use the number of nodes and the total edge weight to measure the abnormal degree, so as to obtain the second abnormal degree of the current shipping address association graph under the second abnormal structure.

[0094] Exemplarily, S450 may include: based on the current number of nodes corresponding to the current shipping address association graph centered on the current target historical shipping address and the variation relationship between the number of nodes and the total edge weight obtained by pre-fitting, determine the fitted total edge weight corresponding to the current number of nodes; based on the current total edge weight and the fitted total edge weight corresponding to the current shipping address association graph, determine the second deviation degree under the second abnormal structure; based on the preset density identification method, the current number of nodes and the current total edge weight corresponding to the current shipping address association graph, determine the second local abnormal degree under the second abnormal structure; based on the second deviation degree and the second local abnormal degree, determine the second abnormal degree under the second abnormal structure.

[0095] Specifically, the number of edges and the total edge weight in the shipping address association graph follow a power-law distribution, that is Take the natural logarithm of the number of edges and the total edge weight respectively, and then perform linear regression fitting to obtain the slope and intercept, that is, ln W = β 1 *lnNA + β 2 . Determine the slope β 1 and intercept β 2 through historical data. After that, the variation relationship between the number of nodes obtained by fitting and the total edge weight can be obtained. Based on this variation relationship and the current shipping address association graph U i corresponding to the current number of nodes NA i , the fitted total edge weight can be determined Based on the current total edge weight W i corresponding to the current shipping address association graph and the fitted total edge weight , determine the maximum and minimum values of the current total edge weight W i and the fitted total edge weight , as well as the difference between the current total edge weight and the fitted total edge weight, and use log to amplify this difference. Multiply the ratio between the maximum and minimum values by the amplified result to obtain the second deviation degree outline2(i) under the second abnormal structure.

[0096] Based on the current number of nodes and the current total edge weight corresponding to the current shipping address association graph, density identification can be performed to determine the second local abnormal degree lof2(i) under the second abnormal structure. The second deviation degree and the second local abnormal degree can be normalized and summed, so as to comprehensively measure the second abnormal degree of the current shipping address association graph U i under the second abnormal structure, that is, the second abnormal degree out-score2(i) corresponding to the current target historical shipping address.

[0097] S460. Determine the third abnormal degree under the third abnormal structure based on the current total edge weight corresponding to the current shipping address association graph and the maximum eigenvalue of the current adjacency matrix.

[0098] Among them, the current adjacency matrix can refer to the weighted adjacency matrix composed of the edge weights in the current shipping address association graph. The maximum eigenvalue of the current adjacency matrix can refer to the maximum eigenvalue of the current adjacency matrix. It should be noted that the third abnormal structure does not conform to the performance of the shipping addresses of normal objects, that is, there is an edge with abnormal weights between the central node and the neighbor nodes. The central node of this structure is very likely to correspond to the goods collection address of the abnormal object.

[0099] Specifically, the third abnormal structure can use the total edge weight and the maximum eigenvalue of the adjacency matrix to measure the abnormal degree, so as to obtain the third abnormal degree of the current shipping address association graph under the third abnormal structure.

[0100] Exemplarily, S460 may include: determining a fitted maximum eigenvalue of the adjacency matrix corresponding to the current total edge weight based on the change relationship between the current total edge weight corresponding to the current shipping address association graph centered on the current target historical shipping address and the total edge weight and the maximum eigenvalue of the adjacency matrix obtained by pre-fitting; determining a third deviation degree under the third abnormal structure based on the current maximum eigenvalue of the adjacency matrix and the fitted maximum eigenvalue of the adjacency matrix corresponding to the current shipping address association graph; determining a third local abnormality degree under the third abnormal structure based on a preset density identification method, the current total edge weight and the current maximum eigenvalue of the adjacency matrix corresponding to the current shipping address association graph; and determining a third abnormal degree under the third abnormal structure based on the third deviation degree and the third local abnormality degree.

[0101] Specifically, the total edge weight and the maximum eigenvalue of the adjacency matrix in the shipping address association graph follow a power-law distribution, that is Taking the natural logarithm of the total edge weight and the maximum eigenvalue of the adjacency matrix respectively, and then performing linear regression fitting to obtain the slope and intercept, that is, lnλ = β 1 *ln W + β 2 . The slope β 1 and the intercept β 2 are determined through historical data, and then the change relationship between the pre-fitted total edge weight and the maximum eigenvalue of the adjacency matrix can be obtained. Based on this change relationship and the current total edge weight W i corresponding to the current shipping address association graph U i , the fitted maximum eigenvalue of the adjacency matrix can be determined Based on the current maximum eigenvalue of the adjacency matrix λ i corresponding to the current shipping address association graph and the fitted maximum eigenvalue of the adjacency matrix determine the maximum and minimum values of the current maximum eigenvalue of the adjacency matrix λ i and the fitted maximum eigenvalue of the adjacency matrix , as well as the difference between the current maximum eigenvalue of the adjacency matrix and the fitted maximum eigenvalue of the adjacency matrix, and use log to amplify this difference, and multiply the ratio between the maximum and minimum values by the amplified result to obtain the third deviation degree outline3(i) under the third abnormal structure.

[0102] Based on the current total edge weight and the current maximum eigenvalue of the adjacency matrix corresponding to the current shipping address association graph for density identification, the third local abnormality degree lof3(i) under the third abnormal structure can be determined. The third deviation degree and the third local abnormality degree can be normalized and summed, so as to comprehensively measure the third abnormal degree of the current shipping address association graph U i under the third abnormal structure, that is, the third abnormal degree out-score3(i) corresponding to the current target historical shipping address.

[0103] S470. Determine the target abnormal degree corresponding to the current target historical shipping address based on the first abnormal degree, the second abnormal degree, and the third abnormal degree.

[0104] Specifically, the first abnormal degree, the second abnormal degree, and the third abnormal degree can be weighted and summed, and the obtained summation result is used as the target abnormal degree corresponding to the current target historical shipping address, so that the target abnormal degree corresponding to each target historical shipping address can be comprehensively measured, further improving the accuracy of address recognition.

[0105] S480. Determine the target shipping address of the normal object served by the abnormal object based on the target abnormal degree corresponding to the target historical shipping address.

[0106] The technical solution of this embodiment determines the first abnormal degree under the first abnormal structure based on the current number of nodes and the current number of edges corresponding to the current shipping address association graph, determines the second abnormal degree under the second abnormal structure based on the current number of nodes and the current total edge weight corresponding to the current shipping address association graph, and determines the third abnormal degree under the third abnormal structure based on the current total edge weight and the maximum eigenvalue of the adjacency matrix corresponding to the current shipping address association graph. Thus, the target abnormal degree corresponding to each target historical shipping address can be comprehensively measured, further improving the accuracy of address recognition.

[0107] The following is an embodiment of the address recognition device provided by the embodiments of the present invention. This device belongs to the same inventive concept as the address recognition methods in the above embodiments. For the details not described in detail in the embodiment of the address recognition device, reference can be made to the embodiments of the above address recognition methods.

[0108] Figure 5 FIG. is a schematic structural diagram of an address recognition device provided by an embodiment of the present invention. This embodiment is applicable to the situation of identifying the shipping address of the normal object served by the abnormal object. As Figure 5 shown, the device specifically includes: a historical shipping task acquisition module 510, a target historical shipping address determination module 520, a shipping address association graph determination module 530, a target abnormal degree determination module 540, and a target shipping address determination module 550.

[0109] Among them, the historical delivery task acquisition module 510 is used to acquire the historical delivery tasks of the abnormal object; the target historical delivery address determination module 520 is used to determine, based on the historical delivery tasks, the target historical delivery addresses that meet the preset stable delivery conditions from the historical delivery addresses of the abnormal object; the delivery address association graph determination module 530 is used to determine, based on the target historical delivery tasks corresponding to the target historical delivery addresses, the delivery address association graphs centered on each of the target historical delivery addresses; the target abnormality degree determination module 540 is used to determine, based on the delivery address association graphs, the target abnormality degrees corresponding to each of the target historical delivery addresses; the target delivery address determination module 550 is used to determine, based on the target abnormality degrees corresponding to the target historical delivery addresses, the target delivery addresses of the normal objects served by the abnormal object.

[0110] The technical solution of this embodiment determines, based on the historical delivery tasks of the abnormal object, the target historical delivery addresses that meet the preset stable delivery conditions from the historical delivery addresses of the abnormal object, and determines, based on the target historical delivery tasks corresponding to the target historical delivery addresses, the delivery address association graphs centered on each target historical delivery address. Based on the delivery address association graphs, the target abnormality degrees corresponding to each target historical delivery address are determined, and based on the target abnormality degrees corresponding to the target historical delivery addresses, the target delivery addresses of the normal objects served by the abnormal object are determined, thereby realizing the automatic recognition of the delivery addresses of the normal objects served by the abnormal object without manual participation, greatly improving the address recognition efficiency, and ensuring the objectivity and stability of the recognition results.

[0111] Optionally, the target historical delivery address determination module 520 is specifically configured to:

[0112] Perform information processing on the historical delivery tasks to determine the delivery feature information corresponding to each historical delivery address of the abnormal object; group and cluster the historical delivery addresses based on the delivery feature information to obtain each clustered historical delivery address group; determine the target historical delivery addresses that meet the preset stable delivery conditions based on the average delivery feature information corresponding to the historical delivery address groups.

[0113] Optionally, the delivery address association graph determination module 530 is specifically configured to:

[0114] Based on the target historical shipping tasks corresponding to the target historical shipping addresses, detect whether each pair of target historical shipping addresses has the same shipping attribute information; take each of the target historical shipping addresses as a node, connect the associated nodes that have the same shipping attribute information as the current node, and connect the other associated nodes that have the same shipping attribute information as the current associated node to obtain a shipping address association graph centered on the current node, and determine the edge weights in the shipping address association graph.

[0115] Optionally, the shipping address association graph determination module 530 includes:

[0116] A historical shipping task acquisition unit, configured to acquire a first historical shipping task corresponding to a first node of the current edge in the shipping address association graph and a second historical shipping task corresponding to a second node, where the shipping address in the first historical shipping task is the target historical shipping address corresponding to the first node, and the shipping address in the second historical shipping task is the target historical shipping address corresponding to the second node;

[0117] A historical shipping task quantity determination unit, configured to determine the quantity of target first historical shipping tasks that have the same shipping attribute information as at least one of the second historical shipping tasks, and determine the quantity of target second historical shipping tasks that have the same shipping attribute information as at least one of the first historical shipping tasks;

[0118] An edge weight determination unit, configured to determine the edge weight corresponding to the current edge based on the quantity of target first historical shipping tasks, the quantity of target second historical shipping tasks, the total quantity of first historical shipping tasks, and the total quantity of second historical shipping tasks.

[0119] Optionally, the edge weight determination unit is specifically configured to:

[0120] Add the quantity of target first historical shipping tasks and the quantity of target second historical shipping tasks to obtain the quantity of associated tasks corresponding to the current edge; add the total quantity of first historical shipping tasks and the total quantity of second historical shipping tasks to obtain the total quantity of shipping tasks corresponding to the current edge; use the ratio between the quantity of associated tasks corresponding to the current edge and the corresponding total quantity of shipping tasks as the edge weight corresponding to the current edge.

[0121] Optionally, the target anomaly degree determination module 540 includes:

[0122] An anomaly degree determination unit, configured to determine the anomaly degree under each anomaly structure based on at least one anomaly structure and the current shipping address association graph centered on the current target historical shipping address;

[0123] A target anomaly degree determination unit, configured to determine a target anomaly degree corresponding to a current target historical shipping address based on the anomaly degree under each anomaly structure.

[0124] Optionally, the anomaly structure includes at least one of: a first anomaly structure in a star shape or a cluster shape, a second anomaly structure with an abnormal total edge weight, and a third anomaly structure with an abnormal edge weight.

[0125] The anomaly degree determination unit includes:

[0126] A first anomaly degree determination subunit, configured to determine a first anomaly degree under the first anomaly structure based on the current number of nodes and the current number of edges corresponding to the current shipping address association graph centered on the current target historical shipping address.

[0127] A second anomaly degree determination subunit, configured to determine a second anomaly degree under the second anomaly structure based on the current number of nodes and the current total edge weight corresponding to the current shipping address association graph.

[0128] A third anomaly degree determination subunit, configured to determine a third anomaly degree under the third anomaly structure based on the current total edge weight and the maximum eigenvalue of the adjacency matrix corresponding to the current shipping address association graph.

[0129] Optionally, the first anomaly degree determination subunit is specifically configured to:

[0130] Based on the current number of nodes corresponding to the current shipping address association graph centered on the current target historical shipping address and the variation relationship between the number of nodes and the number of edges obtained by pre-fitting, determine the fitted number of edges corresponding to the current number of nodes; based on the current number of edges and the fitted number of edges corresponding to the current shipping address association graph, determine a first deviation degree under the first anomaly structure; based on a preset density identification method, the current number of nodes and the current number of edges corresponding to the current shipping address association graph, determine a first local anomaly degree under the first anomaly structure; based on the first deviation degree and the first local anomaly degree, determine the first anomaly degree under the first anomaly structure.

[0131] Optionally, the target shipping address determination module 550 is specifically configured to:

[0132] Based on the target anomaly degree corresponding to the target historical shipping address, perform grouping and clustering on the target historical shipping address to obtain an abnormal shipping address group and a non-abnormal shipping address group; determine the target historical shipping address in the non-abnormal shipping address group as the target shipping address of the normal object served by the abnormal object.

[0133] The address recognition device provided by an embodiment of the present invention can execute the address recognition method provided by any embodiment of the present invention, and has corresponding functional modules and beneficial effects for executing the address recognition method.

[0134] It should be noted that in the above embodiment of the address recognition device, the various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be realized; in addition, the specific names of the functional units are only for the convenience of mutual distinction and do not limit the protection scope of the present invention.

[0135] Figure 6 It is a schematic structural diagram of an electronic device provided by an embodiment of the present invention. Figure 6 It shows a block diagram of an exemplary electronic device 12 suitable for implementing the embodiments of the present invention. Figure 6 The shown electronic device 12 is only an example and should not bring any limitation to the functions and usage scope of the embodiments of the present invention.

[0136] As Figure 6 shown, the electronic device 12 is presented in the form of a general-purpose computing device. The components of the electronic device 12 may include, but are not limited to: one or more processors or processing units 16, a system memory 28, and a bus 18 connecting different system components (including the system memory 28 and the processing unit 16).

[0137] The bus 18 represents one or more of several types of bus structures, including a memory bus or a memory controller, a peripheral bus, a graphics acceleration port, a processor, or a local bus using any of the multiple bus structures. For example, these architectures include, but are not limited to, Industry Standard Architecture (ISA) bus, Micro Channel Architecture (MAC) bus, Enhanced ISA bus, Video Electronics Standards Association (VESA) local bus, and Peripheral Component Interconnect (PCI) bus.

[0138] The electronic device 12 typically includes a variety of computer system-readable media. These media can be any available media accessible by the electronic device 12, including volatile and non-volatile media, removable and non-removable media.

[0139] The system memory 28 may include computer system-readable media in the form of volatile memory, such as random access memory (RAM) 30 and / or cache memory 32. The electronic device 12 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, a storage system 34 may be used to read and write non-removable, non-volatile magnetic media ( Figure 6 not shown, commonly referred to as a "hard disk drive"). Although Figure 6Not shown in the figure, a disk drive for reading and writing a removable non-volatile disk (such as a "floppy disk") and an optical disk drive for reading and writing a removable non-volatile optical disk (such as a CD-ROM, DVD-ROM or other optical medium) can be provided. In these cases, each drive can be connected to the bus 18 through one or more data medium interfaces. The system memory 28 may include at least one program product having a set (such as at least one) of program modules configured to perform the functions of the embodiments of the present invention.

[0140] A program / utility 40 having a set (at least one) of program modules 42 can be stored, for example, in the system memory 28. Such program modules 42 include, but are not limited to, an operating system, one or more application programs, other program modules, and program data. Each or some combination of these examples may include the implementation of a network environment. The program modules 42 generally perform the functions and / or methods in the embodiments described in the present invention.

[0141] The electronic device 12 can also communicate with one or more external devices 14 (such as a keyboard, a pointing device, a display 24, etc.), and can also communicate with one or more devices that enable a user to interact with the electronic device 12, and / or communicate with any device that enables the electronic device 12 to communicate with one or more other computing devices (such as a network card, a modem, etc.). Such communication can be carried out through the input / output (I / O) interface 22. In addition, the electronic device 12 can also communicate with one or more networks (such as a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) through the network adapter 20. As shown in the figure, the network adapter 20 communicates with other modules of the electronic device 12 through the bus 18. It should be understood that although not shown in the figure, other hardware and / or software modules can be used in combination with the electronic device 12, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems, etc.

[0142] The processing unit 16 executes various functional applications and data processing by running the programs stored in the system memory 28. For example, it implements the steps of an address recognition method provided by the embodiments of the present invention. The method includes:

[0143] Obtain the historical shipping tasks of the abnormal object;

[0144] Based on the historical shipping tasks, determine a target historical shipping address that meets the preset stable shipping conditions from the historical shipping addresses of the abnormal object;

[0145] Determine a delivery address association graph centered on each of the target historical delivery addresses based on the target historical delivery tasks corresponding to the target historical delivery addresses;

[0146] Determine the target abnormality degree corresponding to each of the target historical delivery addresses based on the delivery address association graph;

[0147] Determine the target delivery address of the normal object served by the abnormal object based on the target abnormality degree corresponding to the target historical delivery address.

[0148] Of course, those skilled in the art can understand that the processor can also implement the technical solutions of the address recognition method provided in any embodiment of the present invention.

[0149] This embodiment provides a computer-readable storage medium, on which a computer program is stored. When the program is executed by a processor, it implements the steps of the address recognition method provided in any embodiment of the present invention. The method includes:

[0150] Obtain the historical delivery tasks of the abnormal object;

[0151] Based on the historical delivery tasks, determine the target historical delivery addresses that meet the preset stable delivery conditions from the historical delivery addresses of the abnormal object;

[0152] Determine a delivery address association graph centered on each of the target historical delivery addresses based on the target historical delivery tasks corresponding to the target historical delivery addresses;

[0153] Determine the target abnormality degree corresponding to each of the target historical delivery addresses based on the delivery address association graph;

[0154] Determine the target delivery address of the normal object served by the abnormal object based on the target abnormality degree corresponding to the target historical delivery address.

[0155] The computer storage medium of the embodiments of the present invention may adopt any combination of one or more computer-readable media. The computer-readable media may be computer-readable signal media or computer-readable storage media. The computer-readable storage media may be, for example, but not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or components, or any combination of the above. More specific examples (non-exhaustive list) of the computer-readable storage media include: electrical connections with one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the above. In this document, the computer-readable storage media may be any tangible medium that contains or stores a program, and the program can be used by or in conjunction with an instruction execution system, device, or component.

[0156] The computer-readable signal media may include data signals propagated in a baseband or as part of a carrier wave, which carry computer-readable program codes. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The computer-readable signal media may also be any computer-readable media other than the computer-readable storage media, and the computer-readable media may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, device, or component.

[0157] The program codes contained on the computer-readable media may be transmitted by any appropriate media, including but not limited to: wireless, wire, optical fiber cable, RF, etc., or any suitable combination of the above.

[0158] The computer program codes for performing the operations of the present invention may be written in one or more programming languages or combinations thereof. The programming languages include object-oriented programming languages such as Java, Smalltalk, C++, and also include conventional procedural programming languages - such as the "C" language or similar programming languages. The program codes may be executed entirely on the user's computer, partially on the user's computer, executed as an independent software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (for example, by using an Internet service provider to connect through the Internet).

[0159] Those of ordinary skill in the art should understand that the above-mentioned modules or steps of the present invention can be implemented by a general-purpose computing device. They can be concentrated on a single computing device or distributed over a network composed of multiple computing devices. Optionally, they can be implemented with program codes executable by a computer device, so that they can be stored in a storage device and executed by the computing device, or they can be separately fabricated into individual integrated circuit modules, or multiple modules or steps among them can be fabricated into a single integrated circuit module for implementation. Thus, the present invention is not limited to any specific combination of hardware and software.

[0160] Note that the above is only the preferred embodiment of the present invention and the applied technical principles. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein. Various obvious changes, re-adjustments, and substitutions can be made by those skilled in the art without departing from the protection scope of the present invention. Therefore, although the present invention has been described in detail through the above embodiments, the present invention is not limited to the above embodiments. Without departing from the concept of the present invention, it can also include more other equivalent embodiments, and the scope of the present invention is determined by the scope of the appended claims.

Claims

1. An address recognition method, characterized in that, it includes: Obtain the historical shipping tasks of the abnormal object; Based on the historical shipping tasks, determine the target historical shipping addresses that meet the preset stable shipping conditions from the historical shipping addresses of the abnormal object; Based on the target historical shipping tasks corresponding to the target historical shipping addresses, determine the shipping address association graph centered on each target historical shipping address; Based on the shipping address association graph, determine the target abnormal degree corresponding to each target historical shipping address; Based on the target abnormal degree corresponding to the target historical shipping address, determine the target shipping address of the normal object served by the abnormal object.

2. The method according to claim 1, characterized in that, The step of determining the target historical shipping addresses that meet the preset stable shipping conditions from the historical shipping addresses of the abnormal object based on the historical shipping tasks includes: Perform information processing on the historical shipping tasks to determine the shipping feature information corresponding to each historical shipping address of the abnormal object; Based on the shipping feature information, perform grouping and clustering on the historical shipping addresses to obtain each group of historical shipping addresses after clustering; Based on the average shipping feature information corresponding to the group of historical shipping addresses, determine the target historical shipping addresses that meet the preset stable shipping conditions.

3. The method according to claim 1, characterized in that, The step of determining the shipping address association graph centered on each target historical shipping address based on the target historical shipping tasks corresponding to the target historical shipping addresses includes: Based on the target historical shipping tasks corresponding to the target historical shipping addresses, detect whether every two target historical shipping addresses have the same shipping attribute information; Take each target historical shipping address as a node, connect the associated nodes with the same shipping attribute information as the current node, and connect other associated nodes with the same shipping attribute information as the current associated node to obtain the shipping address association graph centered on the current node, and determine the edge weight in the shipping address association graph.

4. The method according to claim 3, characterized in that, The step of determining the edge weight in the shipping address association graph includes: Obtain the first historical shipping task corresponding to the first node and the second historical shipping task corresponding to the second node of the current edge in the shipping address association graph, where the shipping address in the first historical shipping task is the target historical shipping address corresponding to the first node, and the shipping address in the second historical shipping task is the target historical shipping address corresponding to the second node; Determine the number of target first historical shipping tasks with the same shipping attribute information as at least one of the second historical shipping tasks, and determine the number of target second historical shipping tasks with the same shipping attribute information as at least one of the first historical shipping tasks; Based on the number of target first historical shipping tasks, the number of target second historical shipping tasks, the total number of first historical shipping tasks, and the total number of second historical shipping tasks, determine the edge weight corresponding to the current edge.

5. The method according to claim 4, characterized in that, Determining the edge weight corresponding to the current edge based on the target first historical delivery task quantity, the target second historical delivery task quantity, the total quantity of the first historical delivery tasks, and the total quantity of the second historical delivery tasks includes: Adding the target first historical delivery task quantity and the target second historical delivery task quantity to obtain the associated task quantity corresponding to the current edge; Adding the total quantity of the first historical delivery tasks and the total quantity of the second historical delivery tasks to obtain the total quantity of delivery tasks corresponding to the current edge; Taking the ratio between the associated task quantity corresponding to the current edge and the total quantity of delivery tasks as the edge weight corresponding to the current edge.

6. The method according to claim 1, wherein, determining the target abnormality degree corresponding to each target historical delivery address based on the delivery address association graph includes: Determining the abnormality degree under each abnormal structure based on at least one abnormal structure and the current delivery address association graph centered on the current target historical delivery address; Determining the target abnormality degree corresponding to the current target historical delivery address based on the abnormality degree under each abnormal structure.

7. The method according to claim 6, wherein, the abnormal structure includes at least one of: a first abnormal structure in the shape of a star or a cluster, a second abnormal structure with an abnormal total edge weight, and a third abnormal structure with an abnormal edge weight; determining the abnormality degree under each abnormal structure based on at least one abnormal structure and the current delivery address association graph centered on the current target historical delivery address includes: Determining the first abnormality degree under the first abnormal structure based on the current number of nodes and the current number of edges corresponding to the current delivery address association graph centered on the current target historical delivery address; Determining the second abnormality degree under the second abnormal structure based on the current number of nodes and the current total edge weight corresponding to the current delivery address association graph; Determining the third abnormality degree under the third abnormal structure based on the current total edge weight and the maximum eigenvalue of the adjacency matrix corresponding to the current delivery address association graph.

8. The method according to claim 7, wherein, determining the first abnormality degree under the first abnormal structure based on the current number of nodes and the current number of edges corresponding to the current delivery address association graph centered on the current target historical delivery address includes: Determining the fitted number of edges corresponding to the current number of nodes based on the current number of nodes corresponding to the current delivery address association graph centered on the current target historical delivery address and the variation relationship between the number of nodes and the number of edges obtained by pre-fitting; Determining the first deviation degree under the first abnormal structure based on the current number of edges and the fitted number of edges corresponding to the current delivery address association graph; Determining the first local abnormality degree under the first abnormal structure based on a preset density recognition method, the current number of nodes, and the current number of edges corresponding to the current delivery address association graph; Determining the first abnormality degree under the first abnormal structure based on the first deviation degree and the first local abnormality degree.

9. The method according to any one of claims 1-8, wherein, Determining the target delivery address of the normal object served by the abnormal object based on the target abnormality degree corresponding to the target historical delivery address includes: Grouping and clustering the target historical delivery address based on the target abnormality degree corresponding to the target historical delivery address to obtain an abnormal delivery address group and a non-abnormal delivery address group; Determining the target historical delivery address in the non-abnormal delivery address group as the target delivery address of the normal object served by the abnormal object.

10. An address recognition device Characterized in that It includes: A historical delivery task acquisition module, configured to acquire the historical delivery tasks of an abnormal object; A target historical delivery address determination module, configured to determine, based on the historical delivery tasks, the target historical delivery addresses that meet the preset stable delivery conditions from the historical delivery addresses of the abnormal object; A delivery address association graph determination module, configured to determine a delivery address association graph centered on each of the target historical delivery addresses based on the target historical delivery tasks corresponding to the target historical delivery addresses; A target abnormality degree determination module, configured to determine the target abnormality degree corresponding to each of the target historical delivery addresses based on the delivery address association graph; A target delivery address determination module, configured to determine the target delivery address of the normal object served by the abnormal object based on the target abnormality degree corresponding to the target historical delivery address.

11. An electronic device Characterized in that The electronic device includes: One or more processors; A memory, configured to store one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the address recognition method according to any one of claims 1-9.

12. A computer-readable storage medium, on which a computer program is stored Characterized in that When the program is executed by a processor, it implements the address recognition method according to any one of claims 1-9.

Images