Address processing method, device, equipment and storage medium
By performing word segmentation processing and label analysis on the target delivery address, combined with the judgment strategy of the historical delivery address set, the problem of low accuracy of illegal address recognition in the existing technology is solved, and more efficient illegal address recognition and interception is achieved.
Patent Information
- Application Number
- CN202210444711.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-04-26
- Publication Date
- 2025-04-18
- Estimated Expiration
- 2042-04-26
AI Technical Summary
The prior art has low accuracy when identifying illegal addresses, has a large maintenance workload, and cannot effectively distinguish between the address to be confirmed and the illegal address.
By performing word segmentation on the target delivery address, multiple address elements and their tags are obtained, and whether they are illegal addresses are determined based on the tags. If it is impossible to determine, a judgment strategy is adopted, combined with the historical delivery address set, confirm whether it is the address to be confirmed, and output a prompt information for the user to confirm.
It improves the accuracy of illegal address recognition, reduces maintenance workload, and enhances the ability to identify and intercept illegal addresses through the process of confirming the address to be confirmed.
Smart Images

Figure CN114841156B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to communication technology, and in particular to an address processing method, device, equipment and storage medium. Background Art
[0002] When ordering goods through online channels, there are cases where abnormal users use illegal addresses to place malicious orders, such as fake addresses, meaningless junk addresses, and other illegal addresses. Since illegal addresses cannot be delivered normally, it is necessary to identify illegal addresses and intercept their corresponding orders.
[0003] At present, illegal addresses are usually identified and blocked by using illegal address templates for precise matching or regular expression fuzzy matching. However, due to the high randomness of illegal addresses and the possibility of misplacing real detailed addresses with non-matching regions, the existing technology has the problems of large maintenance workload and low interception accuracy. Summary of the invention
[0004] The present application provides an address processing method, device, equipment and storage medium, which are used to solve the problem of low accuracy in illegal address recognition caused by the inability to recognize the address to be confirmed in the prior art.
[0005] In a first aspect, the present application provides an address processing method, comprising:
[0006] The target delivery address of the target order is segmented to obtain M address elements and labels of the M address elements, wherein there are N levels of address elements in the M address elements, and both M and N are integers greater than or equal to 2.
[0007] Determine whether the target delivery address is an illegal address according to the labels of the M address elements.
[0008] If it is impossible to determine whether the target delivery address is an illegal address based on the labels of the M address elements, a first judgment strategy is adopted to determine whether the target delivery address is an address to be confirmed; the first judgment strategy includes: determining whether the target delivery address is an address to be confirmed based on at least one second address element among the M address elements and a set of historical delivery addresses; the delivery addresses in the set of historical delivery addresses all include at least one first address element among the M address elements, and the address level to which the at least one first address element belongs is higher than that of the at least one second address element.
[0009] If it is determined that the target delivery address is an address to be confirmed, a prompt message is output, where the prompt message is used to prompt the user to confirm whether the target delivery address is an illegal address.
[0010] Optionally, determining whether the target delivery address is an address to be confirmed based on at least one second address element among the M address elements and the historical delivery address set includes:
[0011] A second address element is determined from the at least one second address element as a third address element to be matched.
[0012] The third address element is used to perform matching processing with the address elements in the historical delivery address set to obtain a matching result.
[0013] If the matching result indicates a matching failure, the target delivery address is determined to be an address to be confirmed.
[0014] If the matching result indicates a successful match, then based on a second judgment strategy located after the first judgment strategy among the preset Y judgment strategies, it is determined whether the target delivery address is an illegal address; Y is an integer greater than or equal to 3.
[0015] Optionally, the using the third address element to perform matching processing with address elements in the historical delivery address set to obtain a matching result includes:
[0016] The third address element is used to perform an exact matching process with the address elements in the historical delivery address set to obtain an initial matching result.
[0017] If the initial matching result indicates a successful match, the initial matching result is used as the matching result.
[0018] If the initial matching result indicates a matching failure, a correction process is performed on the third address element.
[0019] If the third address element is corrected successfully, the corrected third address element is used to perform an accurate matching process with the address elements in the historical delivery address set to obtain the matching result.
[0020] If the deviation correction of the third address element fails, the third address element is used to perform fuzzy matching processing with the address elements in the historical delivery address set to obtain the matching result.
[0021] Optionally, performing a deflection correction process on the third address element to obtain a deflected third address element includes:
[0022] According to the occurrence frequency of address elements having the same pronunciation and the same address level as the third address element in the delivery addresses of the historical orders, the fourth address element is determined from the address elements having the same address level in the delivery addresses of the historical orders.
[0023] The fourth address element is used as the corrected third address element.
[0024] Optionally, determining a third address element from the at least one second address element comprises:
[0025] It is determined whether there is a second address element of the first address level in the at least one second address element.
[0026] If the second address element of the first address level exists, the second address element of the first address level is used as the third address element.
[0027] If there is no second address element of the first address level, it is determined whether there is a second address element of the second address level in the at least one second address element.
[0028] If there is a second address element of the second address level, the second address element of the second address level is used as the third address element; the second address level is higher than the first address level, and the second address level and the first address level are both lower than the address level to which the at least one first address element belongs.
[0029] If the second address element of the second address level does not exist, it is determined that the match fails.
[0030] Optionally, determining whether the target delivery address is an illegal address according to the labels of the M address elements includes:
[0031] According to the judgment order of at least one third judgment strategy that precedes the first judgment strategy among the Y judgment strategies, each third judgment strategy is used in turn to determine whether the target delivery address is an illegal address according to the labels of the M address elements; each third judgment strategy corresponds to an illegal address type.
[0032] If all the third judgment strategies do not lead to the conclusion that the target delivery address is an illegal address, it is confirmed that it is impossible to determine whether the target delivery address is an illegal address.
[0033] Optionally, the method further comprises:
[0034] If the target delivery address is determined to be an illegal address based on any third judgment strategy, the target order is intercepted and processed according to the interception processing method corresponding to the illegal address type to which the target delivery address belongs.
[0035] Optionally, the word segmentation processing is performed on the target delivery address of the target order to obtain M address elements, and before the labels of the M address elements, the following is further included:
[0036] The target delivery address is input into a preset malicious address recognition model to obtain the recognition probability of the target delivery address.
[0037] If the recognition probability indicates that the target delivery address is not a malicious address, the target delivery address of the target order is segmented to obtain M address elements and labels of the M address elements.
[0038] In a second aspect, the present application provides an address processing device, including:
[0039] A processing module is used to perform word segmentation on the target delivery address of the target order to obtain M address elements and labels of the M address elements, wherein there are N levels of address elements in the M address elements, and both M and N are integers greater than or equal to 2.
[0040] The first determination module is used to determine whether the target delivery address is an illegal address according to the labels of the M address elements.
[0041] The second determination module is used to adopt a first judgment strategy to determine whether the target delivery address is an address to be confirmed if it cannot be determined whether the target delivery address is an illegal address based on the labels of the M address elements. The first judgment strategy includes: determining whether the target delivery address is an address to be confirmed based on at least one second address element among the M address elements and a set of historical delivery addresses; the delivery addresses in the set of historical delivery addresses all include at least one first address element among the M address elements, and the address level to which the at least one first address element belongs is higher than that of the at least one second address element.
[0042] The output module is used to output prompt information if it is determined that the target delivery address is an address to be confirmed, and the prompt information is used to prompt the user to confirm whether the target delivery address is an illegal address.
[0043] In a third aspect, the present application provides an address processing device, comprising: a processor, and a memory communicatively connected to the processor;
[0044] The memory stores computer-executable instructions;
[0045] The processor executes the computer-executable instructions stored in the memory to implement the method as described in any one of the first aspects.
[0046] In a fourth aspect, the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer-executable instructions, and when the computer-executable instructions are executed by a processor, they are used to implement the address processing method as described in any one of the first aspects.
[0047] The address processing method, apparatus, device and storage medium provided in the present application obtain multiple address elements of the target delivery address and labels of the address elements by performing word segmentation processing on the target delivery address, and determine whether the target delivery address is an illegal address based on the address element labels. If it cannot be determined, further confirm whether the target delivery address is an address to be confirmed. Based on the confirmation result, the user is prompted to confirm whether the target delivery address is an illegal address, thereby being able to confirm the address to be confirmed and improving the accuracy of identifying illegal addresses. BRIEF DESCRIPTION OF THE DRAWINGS
[0048] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.
[0049] Figure 1 A schematic diagram of an application scenario of an order system provided in an embodiment of the present application;
[0050] Figure 2 A flowchart of an address processing method provided in an embodiment of the present application;
[0051] Figure 3 A schematic diagram of address levels of an address element provided in an embodiment of the present application;
[0052] Figure 4 A flowchart of another address processing method provided in an embodiment of the present application;
[0053] Figure 5 A flowchart of an address processing method example provided in an embodiment of the present application;
[0054] Figure 6 A schematic diagram of the structure of an address processing device provided in an embodiment of the present application;
[0055] Figure 7 A schematic diagram of the structure of an address processing device provided in an embodiment of the present application.
[0056] The above drawings have shown clear embodiments of the present application, which will be described in more detail later. These drawings and text descriptions are not intended to limit the scope of the present application in any way, but to illustrate the concept of the present application to those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION
[0057] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.
[0058] Explanation of the terms involved in the embodiments of this application:
[0059] Address: refers to the geographical location name obtained by naming the detailed geographical location after the geographical area is divided in reality. The format of the address can be, for example, Room 701, Unit 3, No. 56, D Road, District C, City B, Province A. The format of the address is different depending on the way the geographical area is divided. It should be understood that this application only takes the format of the above address as an example, and does not limit the way the geographical area is divided, and the format of the corresponding address is not limited.
[0060] Address element: Each content included in the address that represents a geographical area, actual location, or room number is an address element. For example, Province A, City B, District C, Road D, Community E, Unit 3, Room 701, etc. are all address elements.
[0061] Address level: refers to the level of the address element in the address determined by the size of the geographical area and / or the containment relationship. For example, in a geographical area division method, the size of the province's geographical area is larger than that of the city, and the province can contain the city, then the address level of the province in the address is different from the address level of the city. The level of the address can be determined according to actual needs, for example, the address level of the province can be set higher than the address level of the city.
[0062] The subsequent embodiments of this application are explained using the geographical area division examples in the above-mentioned glossary as examples.
[0063] At present, when ordering goods through online channels, there are many types of illegal addresses, such as addresses that do not exist in reality, addresses with chaotic address element order, and addresses with address elements that cannot refer to addresses. Since the target orders generated by abnormal orderers using illegal addresses to impersonate normal users are usually unable to be delivered. Therefore, it is necessary to identify and intercept illegal addresses after the orderer places an order and before the goods are delivered.
[0064] In the prior art, the following two methods are usually used to identify and block illegal addresses:
[0065] The first type: Illegal address template matching
[0066] When obtaining the target delivery address of the order, the address is segmented to obtain the multiple characters included in the address. Based on the multiple characters included in the address, an address template corresponding to the address is generated. For example, the address "Room 701, Unit 3, No. 56, D Road, District C, City B, Province A" can be templated as "Room X, Unit X, No. X, D Road, District C, City B, Province A". It is determined whether there is an illegal address template that matches the address template corresponding to the address in the preset illegal address templates. When there is an illegal address template that matches the address template corresponding to the address in the preset illegal address templates, the target delivery address is determined to be an illegal address.
[0067] However, this method can only determine whether the target delivery address format conforms to the preset illegal address template to determine whether the address is an illegal address, but cannot identify whether the address level relationship between the address elements of the address is correct, nor can it identify whether the address is a real address in reality. Therefore, the matching accuracy of this method is low.
[0068] The second type: regular expression fuzzy matching
[0069] Get the target delivery address of the order, parse the address according to the address level, and determine whether the address complies with the preset regular expression matching rules through the relationship between address elements of different levels in the address, such as whether the address level relationship between adjacent address elements is correct. If it does not comply with the matching rules, the target delivery address is considered to be an illegal address.
[0070] However, this method has the problem of address fragmentation, that is, it can only determine whether the address-level relationship between address elements is correct, but cannot determine whether the address points to a real geographical location in reality. Therefore, the matching accuracy of this method is also low.
[0071] In view of this, the present application proposes an address processing method, which performs word segmentation processing on the target delivery address, obtains multiple address elements of the target delivery address and the label of the address element, and determines whether the target delivery address is an illegal address based on the address element label. If it cannot be determined, it further confirms whether the target delivery address is an address to be confirmed. Based on the confirmation result, the user is prompted to confirm whether the target delivery address is an illegal address, so that the address to be confirmed can be confirmed, thereby improving the accuracy of identifying illegal addresses.
[0072] The execution subject of this application may be an order system, which may include a user end for providing order services to subscribers and a service end for processing all orders submitted by subscribers, or a system for processing only all orders submitted by subscribers, which is not limited in this application. The following is an example of possible implementation methods of the order system.
[0073] Figure 1 A schematic diagram of an application scenario of an order system provided in an embodiment of the present application, such as Figure 1 As shown, the order system includes a user end and a server end, wherein the user end includes an ordering page.
[0074] The order system is a system for providing ordering services. On the user side, the subscriber can use the ordering page of the order system to perform online ordering services, for example, order physical goods through the ordering page. Among them, the user side can be, for example, a client installed on the user terminal device. On the service side, the user can obtain the information of the target order through the order system, and then provide the subscriber with the physical goods ordered by the user based on the information of the target order. The information of the target order can include, for example, product information, target delivery address, subscriber's contact information, etc. For example, the subscriber can select the physical goods to be ordered through the ordering page of the order system, and enter the target delivery address, contact information and other information to complete the target order and purchase the corresponding physical goods. After the user obtains the target order information, the physical goods are sent to the subscriber according to the target delivery address in the order information to complete the transaction process.
[0075] Among them, the subscriber can enter the target delivery address through the order page in the order system so that the ordered physical goods can be received. The order page can be composed of an input box that is completely input by the subscriber, or it can be composed of a drop-down selection box and an input box. The embodiment of the present application takes the address input module composed of a drop-down selection box and an input box as an example for explanation. Exemplarily, the target delivery address can be composed of four parts, such as province, city, district and county, and detailed address. The province, city, district and county can be selected through the drop-down box, and the detailed address can be entered by the subscriber through the input box.
[0076] After receiving the target order reported by the user, the server of the order system can call the address processing method by calling the interface to process the target delivery address of the target order to identify the type of the target delivery address and further confirm whether the address is an illegal address.
[0077] Below Figure 1 Taking the application scenario of the order system as an example, the technical solution of the embodiment of the present application is described in detail in combination with a specific embodiment. The subsequent embodiments of the present application are described by taking an address processing scenario as an example. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present application will be described below in combination with the accompanying drawings.
[0078] Figure 2is a flowchart of an address processing method provided in an embodiment of the present application, such as Figure 2 As shown, the method includes:
[0079] S201. Perform word segmentation processing on the target delivery address of the target order to obtain M address elements and labels of the M address elements. There are N levels of address elements in the M address elements, and both M and N are integers greater than or equal to 2.
[0080] The order system obtains the target delivery address of the target order, and uses the address segmentation method in the prior art to segment the target delivery address. For example, the bidirectional long short-term memory (LSTM) algorithm in the prior art can be used to divide the target delivery address into M address elements, and for each address element, a corresponding label is added. The label can be used to identify the level of each address element, or the availability of the address element. The availability can include, for example, whether the address element is a repeated address element, or whether the address element is a noise word unrelated to the address information, or whether the address element is a directional word, etc.
[0081] For example, Figure 3 is an address level diagram of an address element provided by this application, such as Figure 3 As shown, the M address elements after the target delivery address segmentation processing will be divided into 7 address levels. The labels of these 7 address levels and the address elements corresponding to each address level are: Level 1 province [1.prov], Level 2 city [2.city], Level 3 prefecture-level city / district / county [3.dist], Level 4 district / town / village / street [4.town], Level 5 road / village / development zone [5.road], Level 6 point of interest [6.poi], and Level 7 house number [7.house].
[0082] In addition, for the M address elements after the target delivery address is segmented, other features of these address elements can also be marked. For example, repeated address elements can be marked as [-1.redundant] labels, directional words can be marked as [0a.orient] labels, and interference words that are not related to address information can be marked as [0b.other] labels.
[0083] Among them, repeated address elements may be, for example, at least two identical address elements after the target delivery address is segmented. For address elements other than the first identical address element, the label corresponding to the address level will not be marked, but will be marked as the [-1.redundant] label. In addition, users can also set the marking rules of the [-1.redundant] label according to actual needs. For example, it can be set that the first three levels of address elements are not marked as [-1.redundant] labels, so as to avoid addresses such as "Beijing, Beijing" being mistakenly marked as repeated address elements. Directional words may be, for example, words indicating direction such as "east" and "east". Interference words may be, for example, words that are not used to express address information, such as "brother, sister, younger brother, younger sister".
[0084] The subsequent embodiments of the present application are described by taking the above-mentioned marking method of address elements as an example. It should be understood that the marking method of address elements in the present application can be determined according to actual needs and is not limited to the above method.
[0085] S202. Determine whether the target delivery address is an illegal address based on the labels of the M address elements.
[0086] The order system may determine whether the target delivery address is an illegal address based on the address level represented by the labels of the M address elements, or the availability of the address element.
[0087] For example, if the labels of the M address elements contain duplicate address element labels, contain multiple identical labels, or lack labels of necessary address elements, then the target delivery address is an illegal address. If the labels of the M address elements do not contain the above conditions, it is impossible to determine whether the target delivery address is an illegal address. The labels of the above necessary address elements may include, for example, a label [6.poi] for address level six.
[0088] S203: If it is not possible to determine whether the target delivery address is an illegal address based on the labels of the M address elements, the first judgment strategy is adopted to determine whether the target delivery address is an address to be confirmed. If the target delivery address is an address to be confirmed, step S204 is executed.
[0089] The first judgment strategy includes: determining whether the target delivery address is an address to be confirmed based on at least one second address element among the M address elements and a set of historical delivery addresses.
[0090] The historical delivery address set is a set of target delivery addresses corresponding to target orders that have been delivered in the order system. The delivery addresses in the historical delivery address set all include at least one first address element among the above M address elements, and the address level of the at least one first address element is higher than that of the at least one second address element.
[0091] The historical delivery address set can be stored in the storage space of the order system, or in a database that can transmit data with the order system. This application does not limit its storage location. The historical delivery address set can directly store each historical delivery address, and when it needs to be called, the addresses in the historical delivery address set are segmented; it can also be stored in the format of each historical delivery address, the address element corresponding to each address, and the label.
[0092] The above-mentioned target delivery address may be, for example, Building 101, G, District F, E Road, D Street, District C, City B, Province A, and its corresponding address elements and their corresponding address element labels are respectively Province A [1.prov], City B [2.city], District C [3.dist], Street D [4.town], Road E [5.road], District F [6.poi], Building G 101 [7.house]. However, the above-mentioned address elements may contain geographical locations that do not exist in reality. For example, Street D does not exist in District C, City B, Province A. The cause of the above-mentioned problem may be, for example, that when entering the target delivery address, the subscriber entered other streets that do not belong to District C, or entered a street that belongs to District C but the input result of Street D contains typos, or entered an alias or abbreviation of Street D. Therefore, in this case, even if it is impossible to determine whether the target delivery address is an illegal address through the labels of M address elements through the above-mentioned step S202, it is possible to assist in determining whether the target delivery address is an illegal address based on the historical delivery address set.
[0093] Since the historical delivery address set includes at least one of the M address elements, and the address level of the second address element is lower than that of the first address element, it is possible to determine whether the second address element is included in the historical delivery address set based on whether the second address element is included under the first address element in the historical delivery address set, and whether the level relationship between the second address element and the first address element is correct, thereby determining whether the target delivery address is an illegal address. If the second address element is not included under the first address element in the historical delivery address set, it indicates that the second address element is not included in the historical delivery address set, or the level relationship between the second address element and the first address element is incorrect. Therefore, at this time, the target delivery address is an illegal address.
[0094] This application does not limit how to determine whether the target delivery address is a to-be-confirmed address based on at least one second address element among the M address elements and the historical delivery address set. For example, the first second address element and the historical delivery address set may be used for judgment, or multiple second address elements may be used, or the order of the address levels of the second elements may be used for judgment, which may be specifically set according to the needs of the user or the preset judgment rules.
[0095] S204. Output prompt information, where the prompt information is used to prompt the user to confirm whether the target delivery address is an illegal address.
[0096] Because the order system determines that the target delivery address is an illegal address based on the above-mentioned first judgment strategy, the reason why the target delivery address is an illegal address may be that the target delivery address is randomly pieced together by the subscriber using different addresses, or it may be a new legal address that does not exist in the historical delivery address set. Therefore, for the address to be confirmed, the target delivery address needs to be manually confirmed.
[0097] In a possible implementation, the prompt information may be directly output on the order system, for example, the prompt information may be output on a user page in the order system.
[0098] In another possible implementation, the order system may output the prompt information to other user devices, such as computers, mobile devices, etc.
[0099] After the user receives the prompt information output by the order system, the target delivery address of the address to be confirmed is confirmed. If the target delivery address is an illegal address after the user's confirmation, the user can input instructions to the order system, or the order system intercepts the target delivery address according to the judgment result input by the user. If the target delivery address is not an illegal address after the user's confirmation, the user can input instructions to the order system, or the order system further judges the target delivery address according to the judgment result input by the user.
[0100] The address processing method provided in the present application obtains address elements and labels of address elements through word segmentation processing of the target delivery address of the target order. When it is impossible to determine whether the target delivery address is an illegal address, it also determines whether the target delivery address belongs to a to-be-confirmed address through a set of historical delivery addresses, thereby further determining whether the target delivery address is an illegal address based on the actual geographical location, thereby improving the accuracy of illegal address identification.
[0101] The following describes how to use the first judgment strategy to determine whether the target delivery address is an address to be confirmed. That is, how to determine whether the target delivery address is an address to be confirmed based on at least one second address element among the M address elements and the historical delivery address set.
[0102] Figure 4 is a flow chart of another address processing method provided in an embodiment of the present application, such as Figure 4 As shown in a possible implementation manner, the above step S203 may include the following steps:
[0103] S401. Determine a second address element from at least one second address element as a third address element to be matched.
[0104] The third address element is a second address element confirmed according to a preset rule among at least one second address element, and is used for matching with address elements in the historical delivery address set.
[0105] How the order system determines the third address element to be matched from at least one second address element may be implemented in the following ways, for example:
[0106] The first method is to search for a second address element according to the address level of the second address element, from low level to high level, or from high level to low level, or randomly, and determine the second address element as the third address element to be matched.
[0107] For example, determine whether there is a second address element of the first address level in the at least one second address element. If there is a second address element of the first address level, use the second address element of the at least one second address element belonging to the first address level as the third address element. If there is no second address element of the first address level, determine whether there is an address element of the second address level in the at least one second address element. That is to say, when there is no address element of the first address level in the at least one second address element, it is necessary to determine whether there is an address level whose matching priority is second only to the first address level, that is, the second address level.
[0108] If the second address element of the second address level exists, the second address element of the second address level is used as the third address element. If the second address element of the second address level does not exist, it is determined that the match fails.
[0109] Taking the search for the second address element in order from low level to high level as an example, the second address level is higher than the first address level, that is, the lower the address level, the higher the corresponding matching priority. Exemplarily, based on the division of the target delivery address into seven levels in the aforementioned embodiment, the first address level can be set to the sixth-level address element [6.poi], and the second address level can be set to the fifth-level address element [5.road]; or the first address level can be set to the fifth-level address element [5.road], and the second address level can be set to the fourth-level address element [4.town]. That is, under this implementation, the address elements of the lower address level are matched first. If no match is found, the address elements of the higher level are matched, so that the multiple address elements of the target delivery address are matched one by one.
[0110] The second method is to search for multiple second address elements according to the address levels of the second address elements, from low level to high level, or from high level to low level, or randomly, and determine the multiple second address elements as the third address elements to be matched.
[0111] In this implementation, there are multiple third address elements to be matched, and all the third address elements to be matched can be matched in sequence according to a preset matching order.
[0112] The third method is to use all the second address elements as the third address elements to be matched, and to match all the third address elements to be matched in sequence according to a preset matching order.
[0113] Fourth, obtain the target address level, which can be determined according to actual needs. According to the target address level, find a second address element with an address level of the target address level in at least one second address element, and determine the corresponding second address element as the third address element to be matched. There may be multiple corresponding second address elements, or there may be no second address element corresponding to the target address level.
[0114] S402: Use the third address element to perform matching processing with the address elements in the historical delivery address set to obtain a matching result.
[0115] If the third address element is consistent with at least one address element in the historical delivery address set, it means that the third address element is an address element in the historical delivery address set, indicating that the target delivery address is a candidate legal address, and the above matching result is a successful match.
[0116] If the third address element is inconsistent with all the address elements in the historical delivery address set, it means that the third address element is an address element that does not belong to the historical delivery address set, indicating that even if matching is performed through the historical delivery address set, it cannot be concluded that the target delivery address is a candidate legal address, and the matching result is a matching failure.
[0117] The above matching process can be implemented, for example, in the following ways:
[0118] Implementation method 1: first use the third address element to perform an accurate matching process with the address elements in the historical delivery address set to obtain an initial matching result. If the initial matching result indicates a successful match, the initial matching result is used as the matching result; if the initial matching result indicates a failed match, the third address element is corrected, and based on the result of the correction process, the third address element is matched again with the historical delivery address set to obtain the matching result.
[0119] The above-mentioned precise matching process refers to determining whether there is an address element in the historical delivery address set that is completely consistent with the third address element. The completely consistent address element here means that the content of the address element is completely consistent, and the level of the address element is completely consistent. If the third address element is completely consistent with the address element in the historical delivery address set, the initial matching result indicates a successful match; otherwise, the initial matching result indicates a failed match.
[0120] Implementation method 2: first use the third address element to perform deviation correction processing, and then use the corrected third address element to perform precise matching processing with the address elements in the historical delivery address set to obtain a matching result.
[0121] In this implementation, the correction processing method for the third address element and the above-mentioned precise matching method are similar to the method in the aforementioned embodiment 1, and are not described in detail in this application.
[0122] Implementation method three: directly use the third address element to perform precise matching with the address elements in the historical delivery address set to obtain the matching result.
[0123] In this implementation, the above-mentioned precise matching method is similar to the precise matching method in the aforementioned embodiment 1, and is not described in detail in this application.
[0124] S403: If the above matching result indicates a matching failure, the target delivery address is determined to be an address to be confirmed.
[0125] If the matching result obtained by the matching process in step S402 is characterized as a match failure, it means that the third address element is an address element that does not belong to the historical delivery address set, and the user needs to manually determine whether the target delivery address is an illegal address. Therefore, the target address is an address to be confirmed.
[0126] S404. If the above matching result indicates a successful match, determine whether the target delivery address is an illegal address based on a second judgment strategy located after the above first judgment strategy among the preset Y judgment strategies; Y is an integer greater than or equal to 3.
[0127] If the matching result obtained through the matching process in the above step S402 is characterized as a successful match, it means that the third address element is an address element belonging to the historical delivery address set, indicating that the target delivery address is a candidate legal address, that is, the target delivery address is not an address to be confirmed, and it is necessary to further determine whether the target delivery address belongs to an illegal address type.
[0128] Among them, the preset Y judgment strategies may include, for example, a first judgment strategy, a third judgment strategy located before the first judgment strategy for judging the illegal address type of the target delivery address, and a second judgment strategy located after the above-mentioned first judgment strategy for judging the type of the target delivery address. The second judgment strategy may be used to judge the illegal address type of the target delivery address, or it may be used to judge the legal address type of the target address. The third judgment strategy may be, for example, the judgment strategy for determining whether the target delivery address is an illegal address as mentioned in the above-mentioned step S202, which will not be repeated here.
[0129] In a possible implementation, if the target delivery address meets the second judgment strategy, the target delivery address can be determined to be an illegal address; if not, the target delivery address can be considered not to be an illegal address.
[0130] Exemplarily, the second judgment strategy may be, for example, to judge whether the labels of the M address elements include the label [-1.redundant] that indicates the repeated appearance of the address element. The second judgment strategy is used to judge whether the target delivery address is a repeated address type, that is, one or more repeated participles appear in the target delivery address, or the repeated participle is repeated 2 or more times in the target delivery address. The above-mentioned repeated address type belongs to the illegal address type with lower risk.
[0131] Another possible implementation manner is that if the target delivery address meets the second judgment strategy, it can be determined that the target delivery address is not an illegal address.
[0132] Exemplarily, if the target delivery address is judged not to be an illegal address as mentioned above and not to be confirmed according to the labels of the M address elements of the target delivery address, then it can be determined that the target delivery address is not an illegal address. Furthermore, the second judgment strategy can also be used to judge whether the target delivery address is a high-quality address or an unlabeled address.
[0133] The above-mentioned high-quality address refers to a target delivery address that is not an illegal address and whose M address elements' labels have missing address levels acceptable to users. For example, the label levels of the M address elements included in the target delivery address include levels 1-4 and levels 6-7. At this time, it is considered that the target delivery address can uniquely determine the actual location in reality, and the target delivery address is considered to be a high-quality address.
[0134] The above-mentioned unmarked address refers to an address type that is neither an illegal address nor a high-quality address.
[0135] The address processing method provided in the embodiment of the present application identifies whether the target delivery address is a to-be-confirmed address through the M address elements of the target delivery address, the labels of the M address elements, and a set of historical delivery orders, based on a matching method and an address correction method, thereby improving the accuracy of confirming the to-be-confirmed address, and further improving the accuracy and efficiency of identifying illegal addresses.
[0136] The correction process included in the above step S402 may include, for example, the following implementation methods:
[0137] If the above initial matching result indicates a successful match, it means that the third address element is an address element existing in the historical delivery address set, and the target delivery address is a candidate legal address. Therefore, the above initial matching result is used as the matching result, and the matching result is a successful match.
[0138] If the initial matching result indicates a match failure, it means that the third address element is not an address element in the historical delivery address set, or the third address element may contain erroneous content due to input errors. Therefore, it is necessary to further eliminate the influence of the third address element input by the subscriber containing typos and perform correction processing on the third address element.
[0139] A possible implementation method is to perform correction processing through the pronunciation of the third address element. In the case that there may be typos with the same pronunciation in the third address element, the address information represented by the third address element may be included in the historical delivery address set. For example, due to an input error, the subscriber inputs "Heping Home" as "Heping Home", that is, the third address element mentioned above includes "Heping Home", and the third address element is accurately matched with the corresponding historical delivery address set. Since the Chinese characters corresponding to the pronunciation of "he" are different, the match fails. However, the address information represented by the third address element is "Heping Home", which is included in the historical delivery address set.
[0140] In this embodiment, the fourth address element is determined from the address elements with the same address level as the third address element in the delivery address of the historical orders according to the frequency of occurrence of the address elements with the same pronunciation and the same address level as the third address element in the delivery address of the historical orders.
[0141] The pronunciation of the third address element is obtained, and according to the pronunciation of the third address element, all historical delivery address elements with the same pronunciation as the third address element and the same address level are searched in the historical delivery address element correction set. The historical delivery address element correction set includes historical delivery address elements with correct input and historical delivery address elements with deviations caused by the above-mentioned input errors. According to all the historical delivery address elements searched, and the preset rules, it is judged whether the third address element can be calibrated. The preset rule can be, for example, to compare the third address element with the address element with the largest number in the historical delivery address element correction set searched above. If the difference between the third address element and the address element with the largest number is larger, it is considered that the reliability of its correction is higher. Among them, the address element with the largest number in the historical delivery address element correction set is the fourth address element. Further, the user can set the preset conditions according to actual needs. When the difference between the third address element and the address element with the largest number meets the preset conditions, the third address element is corrected; when the difference between the third address element and the address element with the largest number does not meet the preset conditions, the correction result is output as a correction failure.
[0142] For example, the above-mentioned preset conditions may include one or more of the following conditions:
[0143] Condition A: If among all the historical delivery address elements that have the same pronunciation as the third address element and the same address level, the number of deviated historical delivery address elements that are exactly the same as the third address element is less than 2, then the third address element will be corrected to the address element with the largest number in the corrected set of historical delivery address elements searched above.
[0144] Condition B: If the ratio of the number of address elements with the largest number in the searched historical delivery address element correction set and the number of historical delivery address elements with the same deviation as the third address element among all historical delivery address elements with the same pronunciation and address level as the third address element is greater than or equal to n, then the third address element is corrected to the address element with the largest number in the searched historical delivery address element correction set. The above n can be set by the user according to actual needs.
[0145] Condition C: If, among all the historical delivery address elements that have the same pronunciation as the third address element and the same address level, the number of address elements with the largest number in the searched historical delivery address element correction set is greater than or equal to 2n; and the ratio of the number of address elements with the largest number in the address elements of the searched historical orders to the number of historical delivery address elements with exactly the same deviation as the third address element is greater than n / 3, then the third address element will be corrected to the address element with the largest number in the searched historical delivery address element correction set.
[0146] Condition D: If the ratio of the number of address elements with the largest number in the searched correction set of historical delivery address elements to the number of historical delivery address elements that are exactly the same as the third address element among all historical delivery address elements that have the same pronunciation as the third address element and the same address level is greater than n; and the number of historical delivery address elements that are exactly the same as the third address element is less than n, then the third address element will be corrected to the address element with the largest number in the searched correction set of historical delivery address elements.
[0147] Condition E: If, among all the historical delivery address elements that have the same pronunciation as the third address element and the same address level, the number of historical delivery address elements that are exactly the same as the third address element is greater than n; and the number of address elements with the largest number in the searched correction set of historical delivery address elements is greater than 5n; and the ratio of the number of address elements with the largest number in the searched correction set of historical delivery address elements to the number of historical delivery address elements that are exactly the same as the third address element is greater than n / 3, then the third address element shall be corrected to the address element with the largest number in the above-mentioned searched correction set of historical delivery address elements.
[0148] It should be understood that the embodiments of the present application only take the above preset conditions as an example, and do not limit the preset conditions used in actual applications.
[0149] Another possible implementation method is to perform correction processing through the text shape of the third address element. For example, it may be a misspelled word with a similar shape. In the case that there may be misspelled words with similar shapes in the third address element, the address information represented by the third address element may be included in the historical delivery address set. For example, due to input errors, the subscriber inputs "Binshui West Road" included in the historical delivery address set as "Binshui West Road", and the third address element can be corrected through the address elements included in the historical delivery address set.
[0150] The method for correcting the third address element in the present application is only based on the above-mentioned implementation example and is not limited here.
[0151] If the correction is successful, the fourth address element is used as the corrected third address element. The corrected third address element is accurately matched with the address elements in the historical delivery address set again to obtain a matching result, which is a successful match.
[0152] If the deviation correction fails, the third address element is used to perform fuzzy matching with the address elements in the above historical delivery address set to obtain a matching result.
[0153] Regarding the above fuzzy matching processing, the following takes the Chinese address as an example to illustrate the method of fuzzy matching processing:
[0154] S4021. Obtain the content of the third address element.
[0155] The Chinese characters or letter combinations included in the third address element can be identified by text recognition, and the content D of the third address element can be decomposed into i Chinese characters or letter combinations according to the original arrangement order of the Chinese characters or letter combinations in the third address element.
[0156] For example, if the third address element is "finance SOHO", the address element is decomposed into "finance", "finance", and "SOHO".
[0157] S4022. Calculate the importance weight of each Chinese character or letter combination in the historical delivery address set; wherein, the fewer times each Chinese character or letter combination appears in the historical delivery address set, the more scarce the Chinese character or letter combination is, and the greater the importance weight is.
[0158] The calculation formula for the importance weight of each Chinese character or letter combination is:
[0159] D i is the number of times the ith Chinese character or letter combination appears in the historical delivery address set, len(F i) is the number of all addresses in the historical delivery address set.
[0160] The importance weight of each Chinese character or letter combination can be obtained based on the number of times each Chinese character or letter combination appears in the historical delivery address set and the number of all addresses in the historical delivery address set.
[0161] S4023. Calculate the similarity between the content D of the third address element and the addresses included in the historical delivery address set.
[0162] The similarity between the content D of the third address element and the addresses included in the historical delivery address set is calculated by the importance weight of each Chinese character or letter combination and the number of Chinese characters or letter combinations in the content D of the third address element.
[0163] The formula for similarity can be, for example,
[0164] Wherein, n is the number of Chinese characters or letter combinations in the content D of the third address element. i The importance weight of each Chinese character or letter combination.
[0165] Through the above-mentioned fuzzy matching processing method, the similarity between the content D of the third address element and the addresses included in the historical delivery address set can be obtained. If the similarity is greater than or equal to the preset threshold, it is considered that the content of the third address element is similar to the addresses included in the historical delivery address set. For example, the content of the third address element can be an abbreviation of the corresponding address included in the historical delivery address set, or an alias, etc. If the similarity is less than the preset threshold, it is considered that the content of the third address element is not similar to the addresses included in the historical delivery address set. The preset threshold can be determined according to actual needs.
[0166] If the matching result of the above fuzzy matching process is similar, the third address element is used to perform fuzzy matching with the corresponding address element in the above historical delivery address set, and the obtained matching result is a successful match; otherwise, the match fails.
[0167] In addition, with respect to step S202, determining whether the target delivery address is an illegal address according to the labels of the M address elements includes:
[0168] According to the judgment order of at least one third judgment strategy among the above Y judgment strategies that is located before the first judgment strategy, each third judgment strategy is used in turn to determine whether the target delivery address is an illegal address according to the labels of the above M address elements; each third judgment strategy corresponds to an illegal address type.
[0169] The third judgment strategy is used to judge the illegal address type of the target delivery address through the labels of the above-mentioned M address elements. If any third judgment strategy concludes that the target delivery address is an illegal address, then the target delivery address is the illegal address type corresponding to the third judgment strategy. This application does not limit the judgment order of the third judgment strategy.
[0170] For example, the illegal address types may include one or more of the following types: malicious addresses, conflicting addresses, and addresses without detailed information. It should be understood that the embodiments of the present application only use the above illegal address types as examples for illustration, and the types and names of illegal addresses can be adjusted according to actual needs, and the present application does not limit them.
[0171] For malicious address types, for example, the target delivery address filled in by the subscriber includes text descriptions that are suspected to be non-address types. For example, the address element "Brothers Grocery Store" contains text descriptions such as "brothers" that are not address types. However, "Brothers Grocery Store" may be an address type text description, so it is difficult to determine whether the target delivery address is a malicious address. Therefore, for the target delivery address containing text descriptions that are suspected to be non-address types as mentioned above, it is necessary to further match the target delivery address based on the labels of the address elements corresponding to the multiple address levels obtained after the address segmentation process, using the same matching process as described in the aforementioned step S402, so as to determine whether the target delivery address is a malicious address type.
[0172] The third judgment strategy corresponding to this type is: by obtaining the labels of the address elements corresponding to the multiple address levels, the target delivery address is matched as described above, so as to judge whether the target delivery address is a malicious address type. If the match is successful, the target delivery address is not a malicious address type; if the match fails, the target delivery address is a malicious address type.
[0173] For conflicting address types, the first three levels of the target delivery address have one or two or three levels of labels appearing multiple times, such that the delivery process cannot distinguish the unique and clear information of the province, city, district, or county. For example, the target delivery address is "Building K, Dahua Apartment, Jiantao 1st Road, Jiangdong Street, Chengdu City, Sichuan Province, Fuling District, Chongqing City, including the first three levels of address elements "Chongqing [1.prov] Chongqing City [2.city] Fuling District [3.dist]", and the first two levels of address elements "Sichuan Province [1.prov] Chengdu City [2.city]".
[0174] The third judgment strategy corresponding to this type is: judging whether the labels of the M address elements include at least two address element labels of the same address level, and whether the address element labels of the same address level correspond to different address elements.
[0175] For the type of unfilled detailed address, the target delivery address is too simple and does not have the elements that a general address should have. For example, it only fills in the city / district / county, lacks district / town / village / street, and / or, road / village / development zone, and / or, point of interest information, that is, it lacks [4.town], and / or, [5.road], and / or, [6.poi] tags. In this case, the illegal address type is the type of unfilled detailed address.
[0176] The third judgment strategy corresponding to this type is: judging whether one or more of the [4.town], [5.road], and [6.poi] tags are missing in the tags of the M address elements.
[0177] For example, if the target delivery address and its corresponding label are in the form of "Province A [1.prov] City B [2.city] District C [3.dist] No. 703 [7.house]", or "Province A [1.prov] City B [2.city] District C [3.dist] Street D [4.town]", or "Province A [1.prov] City B [2.city] District C [3.dist] Road E [5.road]", then the target delivery address is an unfilled detailed address type.
[0178] Optionally, if all the third judgment strategies do not result in the result that the target delivery address is an illegal address, it is confirmed that it is impossible to determine whether the target delivery address is an illegal address. That is to say, in this example, if the target delivery address does not belong to any of the above illegal address types, it is impossible to confirm whether the target delivery address is an illegal address, and it is necessary to further determine whether the target delivery address is an address to be confirmed.
[0179] Optionally, if the target delivery address is determined to be an illegal address based on any third judgment strategy, the target order is intercepted and processed according to the interception processing method corresponding to the illegal address type to which the target delivery address belongs.
[0180] Among them, for the target order of the target delivery address of the illegal address type with a higher risk level, it is directly intercepted, that is, the target order is considered to be an illegal order, and the order can be cancelled through the order system. For the target order of the target delivery address of the illegal address type with a lower risk level, the target order is sent to the user, and the user manually confirms whether the order is an illegal order. The above risk level can be determined by the actual needs and experience of the user, so as to classify the illegal address type.
[0181] For example, malicious address types, conflicting address types, and address types without detailed information are classified as illegal address types with a higher risk level, so the target orders corresponding to these three types of illegal address types are directly intercepted. The aforementioned duplicate address types and unconfirmed address types are classified as illegal address types with a lower risk level, so for the target orders corresponding to the duplicate address types, the user manually confirms whether the order is an illegal order. As for the high-quality address type and the unmarked address type in the aforementioned step S404, since they are not illegal address types, their corresponding target orders can be considered normal orders, and the normal order delivery process can be executed later.
[0182] A possible implementation method is that after the target delivery address is identified and confirmed, the data of the target delivery address is stored in a database according to the address type to which the target delivery address belongs, and the database is used to store historical delivery address data. The database may store the data of the target delivery address by directly storing the target delivery address, or by storing each address element of the target delivery address as a piece of data, or by simultaneously storing the target delivery address and each address element corresponding thereto. In addition, the database also stores the address type corresponding to the target delivery address, or stores the target delivery address in different databases according to the address type, or in different classified storage spaces in the same database. Furthermore, the order system may output the data of the address type that the user needs to view or analyze based on the user's request, thereby providing a basis for updating and iterating the location processing method used by the order system.
[0183] In addition, in a possible implementation manner, the target delivery address of the target order is segmented to obtain M address elements, and before the labels of the M address elements, the following is further included:
[0184] The target delivery address is input into a preset malicious address recognition model to obtain the malicious address recognition probability of the target delivery address. The malicious address recognition model can, for example, vectorize the target delivery address text in the prior art, and use a deep neural network binary classification model about malicious addresses pre-trained by preset samples to obtain the malicious address recognition probability of the target delivery address. That is, the malicious address recognition model can output the malicious address recognition probability of the target delivery address for the input target delivery address, and the malicious address recognition probability is used to characterize the probability that the target delivery address is a malicious address. For example, for the malicious address recognition model, the malicious address recognition probability is a numerical result between 0 and 1. If the malicious address recognition probability is greater than or equal to 0.9 and less than or equal to 1, the target delivery address of the target order is considered to be a malicious address, and the target order is directly intercepted. If the malicious address recognition probability is less than 0.9, the target delivery address of the target order is considered to be a non-malicious address, or it may be a non-malicious address.
[0185] Optionally, if the malicious address recognition probability indicates that the target delivery address may be a non-malicious address, the target delivery address of the target order is segmented to obtain M address elements and labels of the M address elements. After completing the segmentation process, the next judgment strategy for the target delivery address can be determined based on the malicious address recognition probability. For example, if the malicious address recognition probability is greater than or equal to 0 and less than 0.7, it indicates that the target delivery address is a non-malicious address, and its next judgment strategy may not judge whether the target delivery address is a malicious address; if the malicious address recognition probability is greater than or equal to 0.7 and less than 0.9, it indicates that it is currently impossible to judge whether the target delivery address is a malicious address, and its next judgment strategy needs to further judge whether the target delivery address is a malicious address.
[0186] Optionally, if the malicious address recognition probability indicates that the target delivery address is a non-malicious address, the target delivery address of the target order is segmented to obtain M address elements and labels of the M address elements. After completing the segmentation process, it is not necessary to determine whether the target delivery address is a malicious address.
[0187] For ease of understanding, the above embodiment is described below with an example of a process.
[0188] Figure 5 FIG. 1 is a flow chart of an example of an address processing method provided in an embodiment of the present application. Figure 5 As shown, the example includes:
[0189] S501. According to the malicious address identification model, the target delivery address of the target order is identified, the malicious address identification probability of the target delivery address is obtained, and it is determined whether the malicious address identification probability is greater than or equal to a first preset threshold. If it is less than, execute step S502. If it is greater than or equal to, the process ends.
[0190] At this time, the first preset threshold is 0.9. If the malicious address recognition probability is greater than or equal to the first preset threshold, the target delivery address of the target order is considered to be a malicious address, and step S510 is executed.
[0191] S502: Perform word segmentation on the target delivery address of the target order, and further determine whether the malicious address recognition probability is greater than or equal to a second preset threshold. If greater than or equal to, execute step S503; if less than, execute step S504.
[0192] Through the aforementioned word segmentation processing method, the target delivery address is divided into M address elements, and a corresponding label is added to each address element.
[0193] At this time, the second preset threshold is 0.7. If the malicious address recognition probability is greater than or equal to the second preset threshold, it indicates that it is currently impossible to determine whether the target delivery address is a malicious address.
[0194] If the malicious address recognition probability is less than the second preset threshold, it indicates that the target delivery address is a non-malicious address.
[0195] S503. Determine whether the target delivery address is a malicious address based on the labels of the M address elements of the target delivery address. If not, execute step S504. If so, execute step S510.
[0196] If the target delivery address is an address with a score between 0.7 and 0.9 after being processed by the aforementioned malicious address identification model, indicating that the target delivery address includes text descriptions that are suspected to be non-address-type, it is necessary to perform the above-mentioned matching processing on the target delivery address through the labels of the address elements corresponding to the multiple address levels obtained. If the match fails, the target delivery address is considered to be a malicious address; if the match succeeds, the target delivery address is considered not to be a malicious address, and further judgment is required on the target delivery address.
[0197] S504. Determine whether the target delivery address is a conflicting address based on the labels of the M address elements of the target delivery address. If not, execute step S505. If yes, execute step S510.
[0198] If the labels of the M address elements of the target delivery address include at least two address element labels of the same address level, and the address elements corresponding to the address element labels of the same address level are different, the target delivery address is considered to be a conflicting address; otherwise, the target delivery address is considered not to be a conflicting address, and further judgment of the target delivery address is required.
[0199] S505. Determine whether the target delivery address is a false address based on the M address elements of the target delivery address. If not, execute step S506. If so, execute step S510.
[0200] According to the M address elements of the target delivery address and the historical false delivery address set, it is determined whether the target delivery address is a false address. The historical false delivery address set is the historical false delivery addresses marked as false addresses in the historical data of the order system.
[0201] The M address elements of the target delivery address are accurately matched with all address elements in the historical false delivery address set. If any address element of the target delivery address matches successfully, the target delivery address is considered to be a false address; if all address elements of the target delivery address fail to match, the target delivery address is considered not to be a false address and further judgment is required on the target delivery address.
[0202] S506. According to the labels of the M address elements of the target delivery address, determine whether the target delivery address is an unfilled detailed address. If not, execute step S507; if so, execute step S510.
[0203] If one or more of the tags [4.town], [5.road], and [6.poi] are missing from the labels of the M address elements of the target delivery address, the target delivery address is considered to be an unfilled detailed address; otherwise, the target delivery address is considered not to be an unfilled detailed address, and further judgment of the target delivery address is required.
[0204] S507. Based on the M address elements of the target delivery address and the labels of the M address elements, determine whether the target delivery address is an address to be confirmed. If not, execute step S508; if so, execute step S510.
[0205] The method for determining whether the target delivery address is an address to be confirmed based on the M address elements of the target delivery address and the labels of the M address elements is the same as the method of the aforementioned steps S401-S402 and will not be repeated here.
[0206] S508. Determine whether the target delivery address is a duplicate address based on the labels of the M address elements of the target delivery address. If not, execute step S509; if so, execute step S510.
[0207] If the labels of the M address elements of the target delivery address include the label [-1.redundant] representing the repeated appearance of the address element, the target delivery address is considered to be a duplicate address; if not, the target delivery address is considered not to be a duplicate address and further judgment is required on the target delivery address.
[0208] S509. According to the labels of the M address elements of the target delivery address, determine whether the target delivery address is a high-quality address or an unlabeled address. After the determination is completed, execute step S510.
[0209] If the labels of the M address elements of the target delivery address include labels of all address levels, the target delivery address is considered to be a high-quality address; if not, the target delivery address is considered to be an unlabeled address.
[0210] S510. According to the target delivery address, the target order corresponding to the target delivery address is processed.
[0211] Among them, malicious addresses, conflicting addresses, false addresses, and addresses without detailed information are illegal addresses with a higher risk level. Therefore, the order system will directly intercept and process the target orders corresponding to these four types of illegal addresses.
[0212] Duplicate addresses and addresses to be confirmed are illegal addresses with a relatively low risk level. Therefore, for target orders corresponding to duplicate addresses and addresses to be confirmed, users are required to manually confirm whether the order is an illegal order.
[0213] For the aforementioned high-quality addresses and unmarked addresses, since they do not have the characteristics of the aforementioned illegal addresses, it can be considered that these two types are not illegal addresses, and the corresponding target orders of the two types are regarded as normal orders, and the normal order delivery process can be executed subsequently.
[0214] In the address processing method example, the corresponding relationship of the preset Y judgment strategies is shown in the following Table 1:
[0215] Table 1
[0216]
[0217] The address processing method provided in the embodiment of the present application first uses a malicious address model to perform preliminary screening on the target orders, excludes the target orders corresponding to the target delivery addresses with a higher probability of malicious address recognition, and then obtains multiple address elements and labels of the target delivery address by performing word segmentation processing on the target delivery address, determines the type of the target delivery address according to the address element label, identifies the target delivery address of the illegal address type, and intercepts the target delivery address according to different illegal address types, thereby improving the accuracy and efficiency of identifying and intercepting illegal addresses.
[0218] Figure 6 This is a schematic diagram of the structure of an address processing device provided in an embodiment of the present application. Figure 6 As shown, the address processing device includes: a processing module 11, a first determining module 12, a second determining module 13, and an output module 14. In a possible implementation, it also includes: an intercepting module 15 and an acquiring module 16.
[0219] The processing module 11 is used to perform word segmentation processing on the target delivery address of the target order to obtain M address elements and labels of the M address elements, where there are N levels of address elements in the M address elements, and M and N are both integers greater than or equal to 2.
[0220] The first determination module 12 is used to determine whether the target delivery address is an illegal address according to the labels of the M address elements.
[0221] The second determination module 13 is used to adopt the first judgment strategy to determine whether the target delivery address is an address to be confirmed if it cannot be determined whether the target delivery address is an illegal address based on the labels of the M address elements. The above-mentioned first judgment strategy includes: determining whether the target delivery address is an address to be confirmed based on at least one second address element among the M address elements and a historical delivery address set; the delivery addresses in the historical delivery address set all include at least one first address element among the M address elements, and the address level to which the above-mentioned at least one first address element belongs is higher than that of the above-mentioned at least one second address element.
[0222] The output module 14 is used to output prompt information if it is determined that the target delivery address is an address to be confirmed, and the prompt information is used to prompt the user to confirm whether the target delivery address is an illegal address.
[0223] Optionally, the second determination module 13 is specifically used to determine a second address element from the at least one second address element as a third address element to be matched. The third address element is used to perform matching processing with the address elements in the historical delivery address set to obtain a matching result. If the matching result indicates a matching failure, the target delivery address is determined to be an address to be confirmed. If the matching result indicates a matching success, based on a second judgment strategy that is located after the first judgment strategy among the preset Y judgment strategies, it is determined whether the target delivery address is an illegal address. Y is an integer greater than or equal to 2.
[0224] In a possible implementation, the second determination module 13 is specifically used to use the third address element to perform an exact match process with the address elements in the historical delivery address set to obtain an initial match result. If the initial match result indicates a successful match, the initial match result is used as the match result. If the initial match result indicates a failed match, the third address element is corrected. If the correction of the third address element is successful, the corrected third address element is used to perform an exact match process with the address elements in the historical delivery address set to obtain the match result. If the correction of the third address element fails, the third address element is used to perform a fuzzy match process with the address elements in the historical delivery address set to obtain the match result.
[0225] In this embodiment, the second determination module 13 is specifically used to determine the fourth address element from the address elements with the same address level as the third address element in the delivery address of the historical orders based on the frequency of occurrence of address elements with the same pronunciation and the same address level in the delivery address of the historical orders, and use the fourth address element as the corrected third address element.
[0226] In a possible implementation mode, the second determination module 13 is specifically used to determine whether there is a second address element of the first address level in at least one second address element. If there is a second address element of the first address level, the second address element of the first address level is used as the third address element. If there is no second address element of the first address level, it is determined whether there is a second address element of the second address level in at least one second address element. If there is a second address element of the second address level, the second address element of the second address level is used as the third address element; the second address level is higher than the first address level. If there is no second address element of the second address level, it is determined that the match fails. Both the second address level and the first address level are lower than the address level to which at least one first address element belongs.
[0227] In a possible implementation, the second determination module 13 is specifically used to determine whether the target delivery address is an illegal address according to the labels of M address elements using each third judgment strategy in turn according to the judgment order of at least one third judgment strategy that precedes the first judgment strategy among the Y judgment strategies. If all the third judgment strategies do not produce a result that the target delivery address is an illegal address, it is confirmed that it is impossible to determine whether the target delivery address is an illegal address. Each of the above third judgment strategies corresponds to an illegal address type.
[0228] In this embodiment, optionally, the interception module 15 is used to intercept the target order according to the interception processing method corresponding to the illegal address type to which the target delivery address belongs, if the target delivery address is determined to be an illegal address based on any third judgment strategy.
[0229] Optionally, the acquisition module 16 is used to input the target delivery address of the target order into a preset malicious address recognition model to obtain the recognition probability of the target delivery address before the processing module 11 performs word segmentation processing on the target delivery address to obtain M address elements and labels of the M address elements. If the recognition probability indicates that the target delivery address is not a malicious address, the target delivery address of the target order is subjected to word segmentation processing to obtain M address elements and labels of the M address elements.
[0230] The address processing device provided in the embodiment of the present application can execute the address processing method in the above method embodiment. Its implementation principle and technical effect are similar and will not be repeated here.
[0231] Figure 7 The following is a schematic diagram of the structure of an address processing device provided in an embodiment of the present application. The address processing device is used to execute the aforementioned address processing method, for example, it can be the aforementioned order system. Figure 7 As shown, the address processing device 700 may include: at least one processor 701 , a memory 702 , and in a possible implementation manner, may further include: a communication port 703 .
[0232] The memory 702 is used to store programs. Specifically, the programs may include program codes, and the program codes include computer operation instructions.
[0233] The memory 702 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.
[0234] The processor 701 is used to execute the computer-executable instructions stored in the memory 702 to implement the method described in the above method embodiment. The processor 701 may be a CPU, or an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present application.
[0235] Optionally, the processor 701 may communicate and interact with an external device through the communication port 703 . For example, the external device may be the user device for receiving prompt information as described in the foregoing embodiment.
[0236] In specific implementation, if the communication port 703, the memory 702 and the processor 701 are implemented independently, the communication port 703, the memory 702 and the processor 701 can be connected to each other through a bus and communicate with each other. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus. The bus can be divided into an address bus, a data bus, a control bus, etc., but it does not mean that there is only one bus or one type of bus.
[0237] Optionally, in a specific implementation, if the communication port 703, the memory 702 and the processor 701 are integrated on a chip, the communication port 703, the memory 702 and the processor 701 can communicate through an internal interface.
[0238] The present application also provides a computer-readable storage medium, which may include: a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a disk or an optical disk, and other media that can store program codes. Specifically, the computer-readable storage medium stores program instructions, and the program instructions are used for the methods in the above embodiments.
[0239] The present application also provides a program product, which includes an execution instruction, which is stored in a readable storage medium. At least one processor of the address processing device can read the execution instruction from the readable storage medium, and at least one processor executes the execution instruction so that the address processing device implements the address processing method provided by the above various embodiments.
[0240] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.
Claims
1. An address processing method, characterized in that: include: Perform word segmentation processing on the target delivery address of the target order to obtain M address elements and labels of the M address elements, wherein there are N levels of address elements in the M address elements, and both M and N are integers greater than or equal to 2; Determining whether the target delivery address is an illegal address according to the labels of the M address elements, including: according to the judgment order of at least one third judgment strategy located before the first judgment strategy among the Y judgment strategies, using each third judgment strategy in turn, and determining whether the target delivery address is an illegal address according to the labels of the M address elements; each third judgment strategy corresponds to an illegal address type; the third judgment strategy includes that the labels of the M address elements contain labels with repeated address elements, the labels of the M address elements contain multiple identical labels, or the labels of the M address elements lack labels of necessary address elements; If all the third judgment strategies do not lead to the conclusion that the target delivery address is an illegal address, it is confirmed that it is impossible to determine whether the target delivery address is an illegal address; If it is not possible to determine whether the target delivery address is an illegal address based on the labels of the M address elements, a first judgment strategy is adopted to determine whether the target delivery address is an address to be confirmed; the first judgment strategy includes: determining whether the target delivery address is an address to be confirmed based on whether the first address element in the historical delivery address set includes at least one second address element among the M address elements; the delivery addresses in the historical delivery address set all include at least one first address element among the M address elements, and the address level to which the at least one first address element belongs is higher than that of the at least one second address element; If it is determined that the target delivery address is an address to be confirmed, a prompt message is output, where the prompt message is used to prompt the user to confirm whether the target delivery address is an illegal address.
2. The method according to claim 1, characterized in that The determining whether the target delivery address is an address to be confirmed according to whether the first address element in the historical delivery address set includes at least one second address element in the M address elements includes: Determining a second address element from the at least one second address element as a third address element to be matched; Use the third address element to perform matching processing with the address element under the first address element in the historical delivery address set to obtain a matching result; If the matching result indicates a matching failure, determining the target delivery address as an address to be confirmed; If the matching result indicates a successful match, then based on a second judgment strategy located after the first judgment strategy among the preset Y judgment strategies, it is determined whether the target delivery address is an illegal address; Y is an integer greater than or equal to 3.
3. The method according to claim 2, characterized in that The using the third address element to perform matching processing with the address element under the first address element in the historical delivery address set to obtain a matching result includes: Using the third address element, perform an exact matching process with the address element under the first address element in the historical delivery address set to obtain an initial matching result; If the initial matching result indicates a successful match, the initial matching result is used as the matching result; If the initial matching result indicates a matching failure, performing a correction process on the third address element; If the third address element is successfully corrected, the corrected third address element is used to perform an accurate matching process with the address element under the first address element in the historical delivery address set to obtain the matching result; If the deviation correction of the third address element fails, the third address element is used to perform fuzzy matching processing with the address element under the first address element in the historical delivery address set to obtain the matching result.
4. The method according to claim 3, characterized in that: The step of performing a deflection correction process on the third address element to obtain a deflected third address element includes: According to the occurrence frequency of the address elements having the same pronunciation as the third address element and the same address level in the delivery addresses of the historical orders, determining the fourth address element from the address elements having the same address level in the delivery addresses of the historical orders; The fourth address element is used as the corrected third address element.
5. The method according to any one of claims 2 to 4, characterized in that: The step of determining a third address element from the at least one second address element comprises: determining whether there is a second address element of the first address level in the at least one second address element; If the second address element of the first address level exists, use the second address element of the first address level as the third address element; If there is no second address element of the first address level, determining whether there is a second address element of the second address level in the at least one second address element; If there is a second address element of the second address level, use the second address element of the second address level as the third address element; the second address level is higher than the first address level, and both the second address level and the first address level are lower than the address level to which the at least one first address element belongs; If the second address element of the second address level does not exist, it is determined that the match fails.
6. The method according to claim 1, characterized in that The method further comprises: If the target delivery address is determined to be an illegal address based on any third judgment strategy, the target order is intercepted and processed according to the interception processing method corresponding to the illegal address type to which the target delivery address belongs.
7. The method according to claim 1, characterized in that The target delivery address of the target order is segmented to obtain M address elements, and before the labels of the M address elements, the following is also included: Inputting the target delivery address into a preset malicious address recognition model to obtain the recognition probability of the target delivery address; If the recognition probability indicates that the target delivery address is not a malicious address, the target delivery address of the target order is segmented to obtain M address elements and labels of the M address elements.
8. An address processing device, characterized in that: include: A processing module, configured to perform word segmentation processing on a target delivery address of a target order to obtain M address elements and labels of the M address elements, wherein the M address elements include N levels of address elements, and both M and N are integers greater than or equal to 2; A first determination module is used to determine whether the target delivery address is an illegal address according to the labels of the M address elements, including: according to the judgment order of at least one third judgment strategy located before the first judgment strategy among the Y judgment strategies, each third judgment strategy is used in turn to determine whether the target delivery address is an illegal address according to the labels of the M address elements; each third judgment strategy corresponds to an illegal address type; the third judgment strategy includes labels containing repeated address elements in the labels of the M address elements, the labels of the M address elements contain multiple identical labels, or the labels of the M address elements lack labels of necessary address elements; if the result that the target delivery address is an illegal address is not obtained based on all the third judgment strategies, it is confirmed that it is impossible to determine whether the target delivery address is an illegal address; A second determination module is used for adopting a first judgment strategy to determine whether the target delivery address is an address to be confirmed if it is impossible to determine whether the target delivery address is an illegal address based on the labels of the M address elements; the first judgment strategy includes: determining whether the target delivery address is an address to be confirmed based on whether the first address element in the historical delivery address set includes at least one second address element among the M address elements; the delivery addresses in the historical delivery address set all include at least one first address element among the M address elements, and the address level to which the at least one first address element belongs is higher than that of the at least one second address element; The output module is used to output prompt information if it is determined that the target delivery address is an address to be confirmed, and the prompt information is used to prompt the user to confirm whether the target delivery address is an illegal address.
9. An address processing device, comprising: A processor, and a memory communicatively connected to the processor; The memory stores computer-executable instructions; The processor executes the computer-executable instructions stored in the memory to implement the method according to any one of claims 1 to 7.
10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, and when the computer-executable instructions are executed by a processor, they are used to implement the address processing method according to any one of claims 1 to 7.
11. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the address processing method according to any one of claims 1 to 7 is implemented.
Citation Information
Patent Citations
Method and device for identifying address
CN113076752A
Communication address detection method and device, equipment and medium
CN114372455A