Address aggregation wraparound correction method and apparatus

Abstract

The application relates to the technical field of express delivery, and discloses an address aggregation parcel correction method; the address aggregation parcel correction method comprises the following steps: obtaining a single number and address information of a parcel to be delivered, and performing standardization processing on the address information; based on the standardized address information, initial AOI grouping aggregation of the parcel is completed, an intelligent grouping list is generated according to an aggregation result, and the intelligent grouping list is displayed; single or batch correction operations are performed on the parcel with grouping errors, and a correction sample set is collected; the delivery habit of a delivery man is taken as a reward function, address features and space features are taken as state features, the grouping preference of the delivery man is learned, and the action selection strategy is optimized through the correction sample set; a new parcel order is obtained, and a final intelligent grouping list is generated according to optimal AOI grouping information and displayed; the address aggregation parcel correction method has the advantages that grouping errors caused by non-standard address writing can be solved, and the standardization degree of address analysis can be improved.

Description

Technical Field

[0001] This invention relates to the field of express delivery technology, and in particular to a method and apparatus for address aggregation and parcel correction. Background Technology

[0002] Currently, for each AOI area such as a residential community, shopping mall, or street, delivery personnel sort the packages within their assigned delivery area at the delivery station and then perform a distribution scan. These packages are added to their own pending delivery list, which is then aggregated by AOI group in the intelligent grouping pending delivery list. In actual production, there are instances where addresses are not written correctly, causing packages to not be aggregated into the correct AOI group. There are also packages located at the edge of the AOI area, which, according to delivery personnel's delivery habits, should be assigned to adjacent AOI groups. This affects the delivery personnel's ability to view the number of packages in each group, and subsequent orders will continue to be incorrectly grouped according to the previous grouping logic.

[0003] Therefore, an address aggregation parcel correction method and apparatus are provided to solve the above problems. Summary of the Invention

[0004] The main objective of this invention is to solve the following technical problems in the prior art: improper address writing leads to packages not being aggregated into the correct AOI group; packages located at the edge of the AOI area should be assigned to adjacent AOI groups according to the delivery personnel's delivery habits, which affects the delivery personnel's ability to check the number of packages in each group, and subsequent orders will continue to be incorrectly grouped according to the previous grouping logic.

[0005] The first aspect of the present invention provides a method for address aggregation and packet correction, the method comprising: Obtain the tracking number and recipient address information of the package to be delivered, standardize the recipient address, extract key hierarchical information from the address, and complete and correct non-standard addresses; Based on the standardized address information, spatial correlation analysis is performed on packages at the edge of the AOI area. Address matching degree and spatial distance are used as clustering features to complete the initial AOI grouping and aggregation of packages. A smart grouping list is generated and displayed based on the aggregation results. Perform single or batch correction operations on packages with incorrect grouping, and collect the original AOI grouping information, target AOI grouping information, courier delivery code information and correction time information during the correction operation to form a correction sample set; Using delivery habits as the reward function and address and spatial features as state features, the system learns the deliverymen's grouping preferences and optimizes the action selection strategy by correcting the sample set. Retrieve new package orders, output the optimal AOI grouping, and generate a final intelligent grouping list based on the optimal AOI grouping information for display.

[0006] Optionally, obtaining the tracking number and recipient address information of the package to be delivered, standardizing the recipient address, extracting key hierarchical information from the address, and completing and correcting non-standard addresses include: Obtain the tracking number and associated original recipient address information of the package to be delivered, and collect the tracking number corresponding to the package barcode through the distribution scanning function; Retrieve the recipient address text associated with the tracking number from the express delivery system to complete the collection and associated storage of the raw data; The original recipient address text is input into the pre-trained BERT model. The model performs semantic parsing on the address text and identifies key hierarchical entities such as province, city, district, street, community / shopping mall / school in the address through semantic parsing. Mark the semantic relationships between entities and clarify the hierarchical structure of addresses; The model completes and corrects the identified non-standard addresses. If the model detects that the address has missing information, it completes the address by combining the standardized address data in the corpus. If the model detects that the address has typos or abbreviations, it replaces them with standardized expressions according to the semantic matching results. Output structured, standardized address information, extract key level content, and store it.

[0007] Optionally, the step of performing spatial correlation analysis on packages at the edge of the AOI area based on standardized address information, using address matching degree and spatial distance as clustering features to complete the initial AOI grouping and aggregation of packages, and generating an intelligent grouping list for display based on the aggregation results includes: Convert the standardized address information into spatial latitude and longitude coordinates; Through the geocoding interface, the province, city, district, street, and AOI level information in the address are mapped to the corresponding latitude and longitude data, generating a unique spatial coordinate identifier for each package; Determine the feature dimensions and parameters for DBSCAN spatial clustering, and select address matching similarity and latitude-longitude spatial distance as the two core features for clustering; The DBSCAN algorithm is configured with a neighborhood radius ε and a minimum number of points MinPts. The neighborhood radius is set based on the delivery area, and the minimum number of points is determined based on the number of packages in the AOI area during actual delivery. DBSCAN clustering is then used to filter core points. Each package's coordinates are used as a data point. All data points are traversed, and the number of other data points within the neighborhood radius ε for each data point is counted. If the number is greater than or equal to the minimum number of points MinPts, the data point is marked as a core point. If the number is insufficient but the data point falls within the neighborhood of a core point, it is marked as a boundary point. The rest are considered noise points. Clusters are generated based on the core and boundary points. Adjacent core points are grouped into the same cluster, and boundary points are assigned to the cluster of the nearest core point. Noise points are marked separately. Each cluster corresponds to an initial grouping range for an AOI area. Combining the basic AOI area partitioning rules and clustering results, the initial AOI grouping and aggregation of packages is completed. Packages belonging to the same cluster and the same basic AOI range are grouped into one group, and the group name is named after the corresponding AOI name. At the same time, the grouping results are synchronized to the smart grouping list of the Lanpai APP, and the data is displayed by group. The groups are in a collapsed state by default.

[0008] Optionally, address matching similarity is calculated by comparing the text overlap between the package address and the AOI standard address, and spatial distance is calculated using the Euclidean distance formula based on latitude and longitude coordinates.

[0009] Optionally, the step of performing single or batch correction operations on packages with grouping errors, and collecting the original AOI grouping information, target AOI grouping information, courier delivery code information, and correction time information during the correction operation to form a correction sample set includes: If you find that the package grouping does not match the actual delivery needs when you check the smart grouping list, select a single package or select multiple packages to initiate a correction operation. After the system detects the operation command, it enters the correction interaction process; Perform single or batch correction interactive operations. If it is a single correction, the system retrieves the AOI list of the courier's area of ​​responsibility based on the courier's delivery code and displays it on the APP interface. If it is a batch correction, the system confirms the number of selected packages and displays the AOI selection interface for batch correction. The courier selects the target AOI group from the AOI list. To verify the effectiveness of the correction operation, the system checks whether the courier has bound a delivery code. If not, a prompt will pop up and the operation will be terminated. If the courier has bound a code, the system will confirm whether the target AOI selected by the courier is within its area of ​​responsibility. After verification, the system will perform group adjustment and transfer the package from the original AOI group to the target AOI group. Collect relevant data for the correction operation, extract information such as the original AOI group name and code, the target AOI group name and code, the courier's delivery code, and the timestamp of the correction operation from the system, and associate the corresponding package tracking number with the standardized address to form a single correction record. A correction sample set is constructed, and each record generated by manual correction is organized in a unified format, including feature fields and tag fields. New correction records are continuously added to the sample set and stored in a classified manner according to the time dimension and the courier dimension.

[0010] Optionally, the step of using the delivery habits of couriers as the reward function and address and spatial features as state features to learn the grouping preferences of couriers and optimize the action selection strategy by correcting the sample set includes: The address features, spatial features, and delivery features of the courier are used as state features, and each feature is normalized. Based on the delivery habits of couriers, if the output grouping results are consistent with the target AOI grouping manually corrected by the courier, a positive reward value is set; if the results are alternative delivery groups commonly used by couriers, a medium reward value is set; if the results are completely inconsistent with the delivery habits of couriers, a negative reward value is set, thus clarifying the reward rules corresponding to different grouping decisions. Select a specific AOI group as an action. The action space contains all AOI group options corresponding to the courier's delivery code. Output the grouping decision by selecting different actions. Using the state features in the correction sample set as input and the correction result as label, group actions are selected in the action space, the reward value of each selection is calculated according to the reward function, and the model parameters are updated through backpropagation to optimize the action selection strategy.

[0011] Optionally, by establishing a new binding relationship between the recipient address and the AOI-parcel stack, subsequent order grouping logic can be executed. When the recipient address of a new order is the corrected address, the new order will be automatically calculated into the corrected AOI-parcel stack group. After the courier distributes the parcel, the correct grouping data will be displayed in the courier's pickup and delivery APP - smart grouping list.

[0012] A second aspect of the present invention provides an address aggregation package correction device, the address aggregation package correction device comprising: The address standardization module is used to obtain the tracking number and recipient address information of the package to be delivered, standardize the recipient address, extract key hierarchical information from the address, and complete and correct non-standard addresses. The AOI grouping and aggregation module is used to perform spatial correlation analysis on packages at the edge of the AOI area based on standardized address information. It uses address matching degree and spatial distance as clustering features to complete the initial AOI grouping and aggregation of packages, and generates an intelligent grouping list for display based on the aggregation results. The AOI grouping correction module is used to perform single or batch correction operations on packages with grouping errors. It collects the original AOI grouping information, target AOI grouping information, courier delivery code information, and correction time information during the correction operation to form a correction sample set. The AOI grouping optimization module is used to learn the grouping preferences of couriers by using their delivery habits as the reward function and address and spatial features as state features. It optimizes the action selection strategy by correcting the sample set. The final AOI grouping module is used to obtain new package orders, output the optimal AOI grouping, and generate a final intelligent grouping list for display based on the optimal AOI grouping information.

[0013] A third aspect of the present invention provides an electronic device, the electronic device comprising a memory and at least one processor, wherein the memory stores instructions; The at least one processor invokes the instructions in the memory to cause the electronic device to perform the various steps of the address aggregation and package correction method as described above.

[0014] A fourth aspect of the present invention provides a computer-readable storage medium storing instructions that, when executed by a processor, implement the steps of the address aggregation and packet correction method described above.

[0015] In this invention, packages from the same residential area, shopping mall, or street are aggregated and displayed through an intelligent grouping list. Before delivery, couriers can clearly understand the distribution of packages in their delivery area, facilitating route planning and preventing missed deliveries when passing through a particular area. While the intelligent grouping list groups packages from the same residential area, shopping mall, or school into one group, inaccurate automatic grouping can occur due to varying degrees of address writing standards or addresses located at the edge of grouping areas. This necessitates manual correction by delivery personnel. After correction, when new packages with the same address appear, the system learns and assigns them to the correct group, improving delivery efficiency. This invention solves the grouping error problem caused by non-standard address writing, improves the standardization of address resolution, optimizes the grouping logic for packages at the edge of the AOI area, enhances grouping rationality by incorporating spatial location features, learns the personalized delivery habits of couriers, dynamically updates grouping rules, and reduces the frequency of manual correction. Attached Figure Description

[0016] Figure 1 A flowchart of the address aggregation and parcel correction method provided in an embodiment of the present invention; Figure 2 This is a schematic diagram of the address aggregation and package correction device provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0017] This invention provides a method for address aggregation and package correction, including obtaining the tracking number and recipient address information of the package to be delivered; standardizing the recipient address; extracting key hierarchical information from the address; and completing and correcting non-standard addresses. Based on the standardized address information, spatial correlation analysis is performed on packages at the edge of the AOI area, using address matching degree and spatial distance as clustering features to complete the initial AOI grouping and aggregation of packages. An intelligent grouping list is generated and displayed based on the aggregation results. Single or batch correction operations are performed on packages with incorrect grouping, and the original AOI grouping information, target AOI grouping information, courier delivery code information, and correction time information are collected during the correction operation to form a correction sample set. Using delivery habits as the reward function and address and spatial features as state features, this invention learns delivery personnel's grouping preferences and optimizes action selection strategies by correcting the sample set. It acquires new package orders, outputs the optimal AOI group, and generates a final intelligent grouping list for display based on the optimal AOI grouping information. This invention solves the technical problems in existing technologies, such as packages not being aggregated into the correct AOI group due to non-standard address writing, and packages at the edge of the AOI area that should be assigned to adjacent AOI groups according to delivery personnel's habits. This affects delivery personnel's ability to view the number of packages in groups, and subsequent orders continue to be incorrectly grouped according to the previous grouping logic.

[0018] The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" or "having" and any variations thereof are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0019] For ease of understanding, the specific process of the embodiments of the present invention is described below. Please refer to [link / reference]. Figure 1 The first embodiment of the address aggregation package correction method in this invention includes: Obtain the tracking number and recipient address information of the package to be delivered, standardize the recipient address, extract key hierarchical information from the address, and complete and correct non-standard addresses; Specifically, this includes: Obtain the tracking number and associated original recipient address information of the package to be delivered, and collect the tracking number corresponding to the package barcode through the distribution scanning function; Retrieve the recipient address text associated with the tracking number from the express delivery system to complete the collection and associated storage of the raw data; The original recipient address text is input into the pre-trained BERT model. The model performs semantic parsing on the address text and identifies key hierarchical entities such as province, city, district, street, community / shopping mall / school in the address through semantic parsing. Mark the semantic relationships between entities and clarify the hierarchical structure of addresses; The model completes and corrects the identified non-standard addresses. If the model detects that the address has missing information, it completes the address by combining the standardized address data in the corpus. If the model detects that the address has typos or abbreviations, it replaces them with standardized expressions according to the semantic matching results. It outputs structured, standardized address information, extracts and stores key level content, and can standardize addresses to prevent allocation errors due to address issues.

[0020] Based on the standardized address information, spatial correlation analysis is performed on packages at the edge of the AOI area. Address matching degree and spatial distance are used as clustering features to complete the initial AOI grouping and aggregation of packages. A smart grouping list is generated and displayed based on the aggregation results. Perform single or batch correction operations on packages with incorrect grouping, and collect the original AOI grouping information, target AOI grouping information, courier delivery code information and correction time information during the correction operation to form a correction sample set; Using delivery habits as the reward function and address and spatial features as state features, the system learns the deliverymen's grouping preferences and optimizes the action selection strategy by correcting the sample set. Retrieve new package orders, output the optimal AOI grouping, and generate a final intelligent grouping list based on the optimal AOI grouping information for display.

[0021] Please see Figure 1 The second embodiment of the address aggregation parcel correction method in this invention includes: Obtain the tracking number and recipient address information of the package to be delivered, standardize the recipient address, extract key hierarchical information from the address, and complete and correct non-standard addresses; Specifically, this includes: Obtain the tracking number and associated original recipient address information of the package to be delivered, and collect the tracking number corresponding to the package barcode through the distribution scanning function; Retrieve the recipient address text associated with the tracking number from the express delivery system to complete the collection and associated storage of the raw data; The original recipient address text is input into the pre-trained BERT model. The model performs semantic parsing on the address text and identifies key hierarchical entities such as province, city, district, street, community / shopping mall / school in the address through semantic parsing. Mark the semantic relationships between entities and clarify the hierarchical structure of addresses; The model completes and corrects the identified non-standard addresses. If the model detects that the address has missing information, it completes the address by combining the standardized address data in the corpus. If the model detects that the address has typos or abbreviations, it replaces them with standardized expressions according to the semantic matching results. It outputs structured, standardized address information, extracts and stores key level content, and can standardize addresses to prevent allocation errors due to address issues.

[0022] Based on the standardized address information, spatial correlation analysis is performed on packages at the edge of the AOI area. Address matching degree and spatial distance are used as clustering features to complete the initial AOI grouping and aggregation of packages. A smart grouping list is generated and displayed based on the aggregation results. Specifically, this includes: Convert the standardized address information into spatial latitude and longitude coordinates; Through the geocoding interface, the province, city, district, street, and AOI level information in the address are mapped to the corresponding latitude and longitude data, generating a unique spatial coordinate identifier for each package; The feature dimensions and parameters of DBSCAN spatial clustering were determined, and address matching similarity and latitude-longitude spatial distance were selected as the two core features of clustering. Address matching similarity was calculated by comparing the text overlap between the package address and the AOI standard address, and spatial distance was calculated using the Euclidean distance formula based on latitude-longitude coordinates. The DBSCAN algorithm is configured with a neighborhood radius ε and a minimum number of points MinPts. The neighborhood radius is set based on the delivery area, and the minimum number of points is determined based on the number of packages in the AOI area during actual delivery. DBSCAN clustering is then used to filter core points. Each package's coordinates are used as a data point. All data points are traversed, and the number of other data points within the neighborhood radius ε for each data point is counted. If the number is greater than or equal to the minimum number of points MinPts, the data point is marked as a core point. If the number is insufficient but the data point falls within the neighborhood of a core point, it is marked as a boundary point. The rest are considered noise points. Clusters are generated based on the core and boundary points. Adjacent core points are grouped into the same cluster, and boundary points are assigned to the cluster of the nearest core point. Noise points are marked separately. Each cluster corresponds to an initial grouping range for an AOI area. Combining the basic AOI area partitioning rules and clustering results, the initial AOI grouping and aggregation of packages is completed. Packages belonging to the same cluster and the same basic AOI are grouped together, with the group name named after the corresponding AOI. The grouping results are synchronized to the smart grouping list in the Lanpai APP, and the data is displayed by group. Groups are collapsed by default. Perform single or batch correction operations on packages with incorrect grouping, and collect the original AOI grouping information, target AOI grouping information, courier delivery code information and correction time information during the correction operation to form a correction sample set; Using delivery habits as the reward function and address and spatial features as state features, the system learns the deliverymen's grouping preferences and optimizes the action selection strategy by correcting the sample set. Retrieve new package orders, output the optimal AOI grouping, and generate a final intelligent grouping list based on the optimal AOI grouping information for display.

[0023] Please see Figure 1 The third embodiment of the address aggregation package correction method in this invention includes: Obtain the tracking number and recipient address information of the package to be delivered, standardize the recipient address, extract key hierarchical information from the address, and complete and correct non-standard addresses; Specifically, this includes: Obtain the tracking number and associated original recipient address information of the package to be delivered, and collect the tracking number corresponding to the package barcode through the distribution scanning function; Retrieve the recipient address text associated with the tracking number from the express delivery system to complete the collection and associated storage of the raw data; The original recipient address text is input into the pre-trained BERT model. The model performs semantic parsing on the address text and identifies key hierarchical entities such as province, city, district, street, community / shopping mall / school in the address through semantic parsing. Mark the semantic relationships between entities and clarify the hierarchical structure of addresses; The model completes and corrects the identified non-standard addresses. If the model detects that the address has missing information, it completes the address by combining the standardized address data in the corpus. If the model detects that the address has typos or abbreviations, it replaces them with standardized expressions according to the semantic matching results. It outputs structured, standardized address information, extracts and stores key level content, and can standardize addresses to prevent allocation errors due to address issues.

[0024] Based on the standardized address information, spatial correlation analysis is performed on packages at the edge of the AOI area. Address matching degree and spatial distance are used as clustering features to complete the initial AOI grouping and aggregation of packages. A smart grouping list is generated and displayed based on the aggregation results. Specifically, this includes: Convert the standardized address information into spatial latitude and longitude coordinates; Through the geocoding interface, the province, city, district, street, and AOI level information in the address are mapped to the corresponding latitude and longitude data, generating a unique spatial coordinate identifier for each package; The feature dimensions and parameters of DBSCAN spatial clustering were determined, and address matching similarity and latitude-longitude spatial distance were selected as the two core features of clustering. Address matching similarity was calculated by comparing the text overlap between the package address and the AOI standard address, and spatial distance was calculated using the Euclidean distance formula based on latitude-longitude coordinates. The DBSCAN algorithm is configured with a neighborhood radius ε and a minimum number of points MinPts. The neighborhood radius is set based on the delivery area, and the minimum number of points is determined based on the number of packages in the AOI area during actual delivery. DBSCAN clustering is then used to filter core points. Each package's coordinates are used as a data point. All data points are traversed, and the number of other data points within the neighborhood radius ε for each data point is counted. If the number is greater than or equal to the minimum number of points MinPts, the data point is marked as a core point. If the number is insufficient but the data point falls within the neighborhood of a core point, it is marked as a boundary point. The rest are considered noise points. Clusters are generated based on the core and boundary points. Adjacent core points are grouped into the same cluster, and boundary points are assigned to the cluster of the nearest core point. Noise points are marked separately. Each cluster corresponds to an initial grouping range for an AOI area. Combining the basic AOI area partitioning rules and clustering results, the initial AOI grouping and aggregation of packages is completed. Packages belonging to the same cluster and the same basic AOI are grouped together, with the group name named after the corresponding AOI. The grouping results are synchronized to the smart grouping list in the Lanpai APP, and the data is displayed by group. Groups are collapsed by default. Perform single or batch correction operations on packages with incorrect grouping, and collect the original AOI grouping information, target AOI grouping information, courier delivery code information and correction time information during the correction operation to form a correction sample set; Specifically, this includes: If you find that the package grouping does not match the actual delivery needs when you check the smart grouping list, select a single package or select multiple packages to initiate a correction operation. After the system detects the operation command, it enters the correction interaction process; Perform single or batch correction interactive operations. If it is a single correction, the system retrieves the AOI list of the courier's area of ​​responsibility based on the courier's delivery code and displays it on the APP interface. If it is a batch correction, the system confirms the number of selected packages and displays the AOI selection interface for batch correction. The courier selects the target AOI group from the AOI list. To verify the effectiveness of the correction operation, the system checks whether the courier has bound a delivery code. If not, a prompt will pop up and the operation will be terminated. If the courier has bound a code, the system will confirm whether the target AOI selected by the courier is within its area of ​​responsibility. After verification, the system will perform group adjustment and transfer the package from the original AOI group to the target AOI group. Collect relevant data for the correction operation, extract information such as the original AOI group name and code, the target AOI group name and code, the courier's delivery code, and the timestamp of the correction operation from the system, and associate the corresponding package tracking number with the standardized address to form a single correction record. A correction sample set is constructed, and each record generated by manual correction is organized in a unified format, including feature fields and tag fields. New correction records are continuously added to the sample set and stored in a classified manner according to the time dimension and the courier dimension.

[0025] Using delivery habits as the reward function and address and spatial features as state features, the system learns the deliverymen's grouping preferences and optimizes the action selection strategy by correcting the sample set. Specifically, this includes: Based on the delivery habits of couriers, if the output grouping results are consistent with the target AOI grouping manually corrected by the courier, a positive reward value is set; if the results are alternative delivery groups commonly used by couriers, a medium reward value is set; if the results are completely inconsistent with the delivery habits of couriers, a negative reward value is set, thus clarifying the reward rules corresponding to different grouping decisions. Select a specific AOI group as an action. The action space contains all AOI group options corresponding to the courier's delivery code. Output the grouping decision by selecting different actions. Using the state features in the correction sample set as input and the correction result as label, group actions are selected in the action space, the reward value of each selection is calculated according to the reward function, and the model parameters are updated through backpropagation to optimize the action selection strategy.

[0026] Retrieve new package orders, output the optimal AOI grouping, and generate a final intelligent grouping list based on the optimal AOI grouping information for display.

[0027] By establishing a new binding relationship between the delivery address and the AOI-parcel stack, subsequent order grouping logic can be executed. When the delivery address of a new order is the corrected address, the new order will be automatically calculated into the corrected AOI-parcel stack group. After the courier distributes the parcel, the correct grouping data will be displayed in the courier's pickup and delivery APP - smart grouping list.

[0028] The address aggregation and parcel correction method in the embodiments of the present invention has been described above. The address aggregation and parcel correction device in the embodiments of the present invention will be described below. Please refer to [link to relevant documentation]. Figure 2 The address aggregation package correction device in this embodiment of the invention includes, for the above embodiments: Address standardization module 201 is used to obtain the tracking number and recipient address information of the package to be delivered, standardize the recipient address, extract key hierarchical information from the address, and complete and correct non-standard addresses. The AOI grouping and aggregation module 202 is used to perform spatial correlation analysis on packages at the edge of the AOI area based on standardized address information, and uses address matching degree and spatial distance as clustering features to complete the initial AOI grouping and aggregation of packages. Based on the aggregation results, an intelligent grouping list is generated for display. The AOI grouping correction module 203 is used to perform single or batch correction operations on packages with grouping errors. It collects the original AOI grouping information, target AOI grouping information, courier delivery code information and correction time information during the correction operation to form a correction sample set. The AOI grouping optimization module 204 is used to learn the grouping preferences of couriers by using their delivery habits as the reward function and address features and spatial features as state features, and optimize the action selection strategy by correcting the sample set. The final AOI grouping module 205 is used to obtain new package orders, output the optimal AOI grouping, and generate a final intelligent grouping list for display based on the optimal AOI grouping information.

[0029] above Figure 2 The address aggregation and package correction device in the embodiments of the present invention will be described in detail from the perspective of modular functional entities. The electronic device in the embodiments of the present invention will be described in detail from the perspective of hardware processing.

[0030] Figure 3 This is a schematic diagram of the structure of an electronic device 700 provided in an embodiment of the present invention. The electronic device 700 can vary significantly due to different configurations or performance characteristics. It may include one or more processors 710 (e.g., one or more processors) and a memory 720, and one or more storage media 730 (e.g., one or more storage devices, including RAM, FLASH, etc.) for storing application programs 733 or data 732. The memory 720 and storage media 730 can be temporary or persistent storage. The program stored in the storage media 730 may include one or more modules (not shown in the diagram), each module including a series of instruction operations on the electronic device 700. Furthermore, the processor 710 may be configured to communicate with the storage media 730 and execute the series of instruction operations in the storage media 730 on the electronic device 700.

[0031] The electronic device 700 may also include one or more power supplies 740, one or more input / output interfaces 750, and / or one or more operating systems 731, such as FreeRTOS, Android, etc. Those skilled in the art will understand that... Figure 3 The illustrated electronic device structure does not constitute a limitation on electronic devices and may include more or fewer components than illustrated, or combine certain components, or have different component arrangements.

[0032] The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein the computer-readable storage medium stores instructions that, when executed on a computer, cause the computer to perform the steps of the address aggregation package correction method.

[0033] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the system, device, or unit described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0034] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, mobile device, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0035] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for address aggregation parcel correction, characterized in that, The address aggregation parcel correction method includes: Obtain the tracking number and recipient address information of the package to be delivered, standardize the recipient address, extract key hierarchical information from the address, and complete and correct non-standard addresses; Based on the standardized address information, spatial correlation analysis is performed on packages at the edge of the AOI area. Address matching degree and spatial distance are used as clustering features to complete the initial AOI grouping and aggregation of packages. A smart grouping list is generated and displayed based on the aggregation results. Perform single or batch correction operations on packages with incorrect grouping, and collect the original AOI grouping information, target AOI grouping information, courier delivery code information and correction time information during the correction operation to form a correction sample set; Using delivery habits as the reward function and address and spatial features as state features, the system learns the deliverymen's grouping preferences and optimizes the action selection strategy by correcting the sample set. Retrieve new package orders, output the optimal AOI grouping, and generate a final intelligent grouping list based on the optimal AOI grouping information for display.

2. The address aggregation and parcel correction method according to claim 1, characterized in that, The process of obtaining the tracking number and recipient address information of the package to be delivered, standardizing the recipient address, extracting key hierarchical information from the address, and completing and correcting non-standard addresses includes: Obtain the tracking number and associated original recipient address information of the package to be delivered, and collect the tracking number corresponding to the package barcode through the distribution scanning function; Retrieve the recipient address text associated with the tracking number from the express delivery system to complete the collection and associated storage of the raw data; The original recipient address text is input into the pre-trained BERT model. The model performs semantic parsing on the address text and identifies key hierarchical entities such as province, city, district, street, community / shopping mall / school in the address through semantic parsing. Mark the semantic relationships between entities and clarify the hierarchical structure of the address; The model completes and corrects the identified non-standard addresses. If the model detects that the address has missing information, it completes the address by combining the standardized address data in the corpus. If the model detects that the address has typos or abbreviations, it replaces them with standardized expressions according to the semantic matching results. Output structured, standardized address information, extract key level content, and store it.

3. The address aggregation and parcel correction method according to claim 1, characterized in that, Based on the standardized address information, spatial correlation analysis is performed on packages at the edge of the AOI area. Address matching degree and spatial distance are used as clustering features to complete the initial AOI grouping and aggregation of packages. The intelligent grouping list is generated and displayed based on the aggregation results, including: Convert the standardized address information into spatial latitude and longitude coordinates; Through the geocoding interface, the province, city, district, street, and AOI level information in the address are mapped to the corresponding latitude and longitude data, generating a unique spatial coordinate identifier for each package; Determine the feature dimensions and parameters for DBSCAN spatial clustering, and select address matching similarity and latitude-longitude spatial distance as the two core features for clustering; The DBSCAN algorithm is configured with a neighborhood radius ε and a minimum number of points MinPts. The neighborhood radius is set based on the delivery area, and the minimum number of points is determined based on the number of packages in the AOI area during actual delivery. DBSCAN clustering is then used to filter core points. Each package's coordinates are used as a data point. All data points are traversed, and the number of other data points within the neighborhood radius ε for each data point is counted. If the number is greater than or equal to the minimum number of points MinPts, the data point is marked as a core point. If the number is insufficient but the data point falls within the neighborhood of a core point, it is marked as a boundary point. The rest are considered noise points. Clusters are generated based on the core and boundary points. Adjacent core points are grouped into the same cluster, and boundary points are assigned to the cluster of the nearest core point. Noise points are marked separately. Each cluster corresponds to an initial grouping range for an AOI area. Combining the basic AOI area partitioning rules and clustering results, the initial AOI grouping and aggregation of packages is completed. Packages belonging to the same cluster and the same basic AOI range are grouped into one group, and the group name is named after the corresponding AOI name. At the same time, the grouping results are synchronized to the smart grouping list of the Lanpai APP, and the data is displayed by group. The groups are in a collapsed state by default.

4. The address aggregation and parcel correction method according to claim 3, characterized in that, Address matching similarity is calculated by comparing the text overlap between the package address and the AOI standard address. Spatial distance is calculated using the Euclidean distance formula based on latitude and longitude coordinates.

5. The address aggregation and parcel correction method according to claim 1, characterized in that, The process of performing single or batch correction operations on packages with grouping errors involves collecting the original AOI grouping information, target AOI grouping information, courier delivery code information, and correction time information during the correction operation to form a correction sample set, including: If you find that the package grouping does not match the actual delivery needs when you check the smart grouping list, select a single package or select multiple packages to initiate a correction operation. After the system detects the operation command, it enters the correction interaction process; Perform single or batch correction interactive operations. If it is a single correction, the system retrieves the AOI list of the courier's area of ​​responsibility based on the courier's delivery code and displays it on the APP interface. If it is a batch correction, the system confirms the number of selected packages and displays the AOI selection interface for batch correction. The courier selects the target AOI group from the AOI list. To verify the effectiveness of the correction operation, the system checks whether the courier has bound a delivery code. If not, a prompt will pop up and the operation will be terminated. If the courier has bound a code, the system will confirm whether the target AOI selected by the courier is within its area of ​​responsibility. After verification, the system will perform group adjustment and transfer the package from the original AOI group to the target AOI group. Collect relevant data for the correction operation, extract information such as the original AOI group name and code, the target AOI group name and code, the courier's delivery code, and the timestamp of the correction operation from the system, and associate the corresponding package tracking number with the standardized address to form a single correction record. A correction sample set is constructed, and each record generated by manual correction is organized in a unified format, including feature fields and tag fields. New correction records are continuously added to the sample set and stored in a classified manner according to the time dimension and the courier dimension.

6. The address aggregation and parcel correction method according to claim 1, characterized in that, The method of using delivery habits as the reward function and address and spatial features as state features to learn delivery personnel's grouping preferences and optimize action selection strategies by correcting the sample set includes: The address features, spatial features, and delivery features of the courier are used as state features, and each feature is normalized. Based on the delivery habits of couriers, if the output grouping results are consistent with the target AOI grouping manually corrected by the courier, a positive reward value is set; if the results are alternative delivery groups commonly used by couriers, a medium reward value is set; if the results are completely inconsistent with the delivery habits of couriers, a negative reward value is set, thus clarifying the reward rules corresponding to different grouping decisions. Select a specific AOI group as an action. The action space contains all AOI group options corresponding to the courier's delivery code. Output the grouping decision by selecting different actions. Using the state features in the correction sample set as input and the correction result as label, group actions are selected in the action space, the reward value of each selection is calculated according to the reward function, and the model parameters are updated through backpropagation to optimize the action selection strategy.

7. The address aggregation and parcel correction method according to claim 1, characterized in that, The process of obtaining new package orders, outputting the optimal AOI grouping, and generating a final intelligent grouping list based on the optimal AOI grouping information for display includes: Obtain the new binding relationship between the delivery address and the AOI-parcel stack, execute the subsequent order grouping logic, and when the delivery address of the new order is the corrected address, the new order is automatically calculated into the corrected AOI-parcel stack group, and a smart group list is generated and displayed based on the corrected group.

8. An address aggregation package correction device, used in the address aggregation package correction method as described in any one of claims 1-7, characterized in that, include: The address standardization module is used to obtain the tracking number and recipient address information of the package to be delivered, standardize the recipient address, extract key hierarchical information from the address, and complete and correct non-standard addresses. The AOI grouping and aggregation module is used to perform spatial correlation analysis on packages at the edge of the AOI area based on standardized address information. It uses address matching degree and spatial distance as clustering features to complete the initial AOI grouping and aggregation of packages, and generates an intelligent grouping list for display based on the aggregation results. The AOI grouping correction module is used to perform single or batch correction operations on packages with grouping errors. It collects the original AOI grouping information, target AOI grouping information, courier delivery code information, and correction time information during the correction operation to form a correction sample set. The AOI grouping optimization module is used to learn the grouping preferences of couriers by using their delivery habits as the reward function and address and spatial features as state features. It optimizes the action selection strategy by correcting the sample set. The final AOI grouping module is used to obtain new package orders, output the optimal AOI grouping, and generate a final intelligent grouping list for display based on the optimal AOI grouping information.

9. An electronic device comprising a memory and at least one processor, wherein the memory stores instructions; characterized in that, The at least one processor invokes the instructions in the memory to cause the electronic device to perform the steps of the address aggregation and parcel correction method as described in any one of claims 1-7.

10. A computer-readable storage medium storing instructions thereon, characterized in that, When the instructions are executed by the processor, they implement the steps of the address aggregation and parcel correction method as described in any one of claims 1-7.

Images