Letter address classification method

Abstract

The invention belongs to the technical field of financial services, and discloses a letter address classification method. The method comprises the following steps: performing address classification on to-be-classified addresses according to a preset classification strategy to obtain first classified addresses; performing detailed address classification on the first classification address according to a preset natural language processing mode to obtain a second classification address; determining an address detail degree corresponding to the second classification address according to the second classification address; and carrying out detailed address classification on the second classification address according to the address detailed degree to obtain a target classification address. Through hierarchical grouping from macroscopic to microscopic, different address types are classified and graded, including classification of a plurality of levels such as regional distribution, administrative regions, provinces, city scales, city levels, city categories, regions, urban and rural attributes and detailed address types, so that the accuracy of letter address classification is improved; and the letter delivery cost is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of financial services, and particularly to a method for classifying correspondence addresses. Background Art

[0002] The handling of debt disputes between financial institutions and debtors has always been a core function in the financial industry. Usually, financial institutions will entrust law firms to issue legal notice letters to urge debtors to comply with contractual obligations or resolve disputes. However, in this area, there are still some important challenges, mainly including the following aspects: Delivery rate issues for diverse address types: When sending legal notice letters, various types of addresses are often involved (Xiongfei Ma, University of Electronic Science and Technology, 2019, pages 1 - 78), such as addresses detailed to street and house numbers (No. XX, Unit XX, Building XX, XX Park, XX District, XX City, XX Province), addresses in rural areas (No. XX, Group XX, XX Village, XX Town, XX District, XX City, XX Province), factory addresses (such as "No. XX Factory, XX Street, XX City, XX Province"), collective household addresses (such as "XX Collective Household, XX District, XX City, XX Province"), and road addresses (such as "No. XX Road, XX Town, XX City, XX Province"). How to effectively judge the delivery status of different address types and how to classify them effectively is an issue that has not been studied in detail. This has led to uncertainty in the delivery effect of legal letters, increasing the time and additional costs for resolving debt disputes; Regional differences and debtor characteristics: There are obvious regional differences between different provinces and regions (Xiao Chen, Yunnan University of Finance and Economics, 2023, pages 1 - 92), including urban debtors, rural debtors, and debtors in county urban areas. These regional characteristics affect the way debtors receive legal notice letters and their willingness to respond. In addition, the laws, regulations, and regional economic conditions in different provinces also affect debtors' willingness and ways of repaying debts, but the research in this regard is not sufficient. Therefore, how to solve the low accuracy of the method for classifying correspondence addresses has become an urgent problem to be solved.

[0003] The above content is only used to assist in understanding the technical solution of the present invention and does not represent an admission that the above content is prior art. Summary of the Invention

[0004] The main purpose of the present invention is to provide a method for classifying correspondence addresses, aiming to solve the technical problem of low accuracy of the existing method for classifying correspondence addresses.

[0005] To achieve the above purpose, the present invention provides a method for classifying correspondence addresses, and the method includes the following steps:

[0006] Classify the address to be classified according to a preset classification strategy to obtain a first classified address;

[0007] Classify the first classified address in detail according to a preset natural language processing method to obtain a second classified address;

[0008] Determine the address detail level corresponding to the second classified address according to the second classified address;

[0009] Classify the second classified address in detail according to the address detail level to obtain a target classified address.

[0010] Optionally, the step of classifying the first classified address in detail according to a preset natural language processing method to obtain a second classified address includes:

[0011] Identify address keywords from the first classified address according to preset keywords to obtain address keywords;

[0012] Determine the keyword type corresponding to the address keyword according to the address keyword;

[0013] Classify the first classified address in detail according to the keyword type corresponding to the address keyword to obtain a second classified address.

[0014] Optionally, the step of classifying the first classified address in detail according to the keyword type corresponding to the address keyword to obtain a second classified address includes:

[0015] When the address keyword is "school", determine that the keyword type is school type;

[0016] Mark the first classified address according to the school type to obtain an initial school address;

[0017] Determine that the initial school address is the second classified address.

[0018] Optionally, the step of determining the address detail level corresponding to the second classified address according to the second classified address includes:

[0019] Determine a first detail level according to the address keyword corresponding to the second classified address;

[0020] Make a detailed level judgment on the second classified address according to the address keyword and the first detail level to obtain a second detail level;

[0021] Determine that the second detail level is the address detail level corresponding to the second classified address.

[0022] Optionally, the step of classifying the second classified address in detail according to the address detail level to obtain a target classified address includes:

[0023] Determine the classification keywords of the second classification address according to the preset classification rules;

[0024] Determine the classification address type of the second classification address according to the classification keywords of the second classification address;

[0025] Mark the second classification address according to the classification address type to obtain the target classification address.

[0026] Optionally, the marking the second classification address according to the classification address type to obtain the target classification address includes:

[0027] When the classification keyword is a warehouse keyword, determine that the classification address type is a place type;

[0028] Mark the second classification address according to the place type to obtain the place type address;

[0029] Determine that the place type address is the target classification address.

[0030] Optionally, the performing address classification on the to-be-classified address according to the preset classification strategy to obtain the first classification address includes:

[0031] Obtain administrative division information, economic zone division information, urban planning information, and township division information;

[0032] Perform address classification on the to-be-classified address according to the administrative division information, the economic zone division information, the urban planning information, and the township division information to obtain the first classification address.

[0033] In the present invention, address classification is performed on the to-be-classified address according to a preset classification strategy to obtain a first classification address; detailed address classification is performed on the first classification address according to a preset natural language processing method to obtain a second classification address; the address detail level corresponding to the second classification address is determined according to the second classification address; detailed address classification is performed on the second classification address according to the address detail level to obtain a target classification address. Through hierarchical grouping from macro to micro, classification and grading of different address types are carried out, including classification at multiple levels such as regional distribution, administrative regions, provinces, city sizes, city levels, city categories, regions, urban-rural attributes, and detailed address types, achieving the improvement of the accuracy of letter address classification and the reduction of the cost of letter delivery. Brief Description of the Drawings

[0034] Figure 1 It is a schematic flowchart of the first embodiment of the letter address classification method of the present invention;

[0035] Figure 2 It is a schematic diagram of letter address classification and grading of an embodiment of the letter address classification method of the present invention;

[0036] Figure 3 This is a schematic flowchart of the second embodiment of the method for classifying correspondence addresses according to the present invention.

[0037] The implementation, functional features, and advantages of the present invention will be further described with reference to the embodiments and the accompanying drawings. Detailed implementation manners

[0038] It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0039] An embodiment of the present invention provides a method for classifying correspondence addresses. Referring to Figure 1 , Figure 1 This is a schematic flowchart of the first embodiment of a method for classifying correspondence addresses according to the present invention.

[0040] In this embodiment, the method for classifying correspondence addresses includes the following steps:

[0041] Step S10: Classify the address to be classified according to a preset classification strategy to obtain a first classified address.

[0042] It can be understood that the preset classification strategy refers to a pre-set address classification strategy. The address classification strategy includes, but is not limited to, economic zone division strategy, administrative region division strategy, urban planning division strategy, and rural / urban / town division strategy, etc. The address to be classified refers to the correspondence address to be classified. The correspondence address includes, but is not limited to, legal correspondence address, logistics correspondence address, etc. The first classified address refers to the address to be classified after being classified by the pre-trial classification strategy.

[0043] In specific implementation, classify the economic zone, administrative region, urban strategy, and rural / urban / town, etc. of the address to be classified according to the pre-set address classification strategy to obtain the classified address to be classified. For example: the address to be classified is No. XX, XX Group, XX Village, XX Town, XX District, XX City, XX Province. Classify the economic zone, administrative region, urban strategy, and rural / urban / town, etc. where the address to be classified is located.

[0044] Step S20: Classify the first classified address in detail according to a preset natural language processing method to obtain a second classified address.

[0045] It can be understood that the preset natural language processing method refers to a pre-set natural language processing method (Natural Language Processing, NLP). The second classified address refers to the first classified address after detailed classification. Classify the first classified address in detail according to the pre-set natural language processing method to obtain the second classified address.

[0046] Step S30: Determine the address detail level corresponding to the second classification address according to the second classification address.

[0047] It can be understood that the address detail level refers to the classification level of the correspondence address. For example, if the address contains keywords such as "community" or "neighborhood", it can be determined that the address is detailed to the community level. Then, after including the above detailed keywords, the system further determines whether there is a house number to distinguish whether it is detailed to the level with a number.

[0048] Step S40: Perform detailed address classification on the second classification address according to the address detail level to obtain the target classification address.

[0049] It can be understood that the target classification address refers to the address to be classified after detailed classification. The second classification address is further classified in detail according to the address detail level to obtain the final classification address.

[0050] It should be noted that in order to accurately obtain the target classification address, further, the performing detailed address classification on the second classification address according to the address detail level to obtain the target classification address includes: determining the classification keywords of the second classification address according to the preset classification rules; determining the classification address type of the second classification address according to the classification keywords of the second classification address; marking the second classification address according to the classification address type to obtain the target classification address.

[0051] It can be understood that the preset classification rules refer to the pre-set ways of classifying addresses according to keyword types. For example, school addresses, farm / forest farm / fishing ground / ranch addresses, and venue addresses, etc. The classification keywords refer to the keywords in the second classification address, and the keywords include but are not limited to schools, farms, and warehouses, etc. The classification address type refers to the type corresponding to the classification keywords. For example, the school keyword corresponds to the school address type, and the warehouse keyword corresponds to the venue address type.

[0052] In specific implementation, the keywords in the second classification address are screened according to the pre-set classification rules to obtain the keywords corresponding to the second classification address, the keyword type corresponding to the second classification address is determined according to the keywords, and finally the second classification address is marked according to the keyword type to obtain the target classification address.

[0053] It should be noted that in order to accurately obtain the target classification address, further, the marking the second classification address according to the classification address type to obtain the target classification address includes: when the classification keyword is a warehouse keyword, determining that the classification address type is the venue type; marking the second classification address according to the venue type to obtain the venue type address; determining the venue type address as the target classification address.

[0054] It can be understood that the second classification address contains warehouse keywords, and then the address type corresponding to the warehouse keywords is determined. Finally, the second classification address is marked as the place address type, that is, the target classification address is obtained.

[0055] It should be noted that, as Figure 2 shown, first, data collection and collation of national administrative divisions are carried out, and three-level, six-level, and seven-level classifications are formed according to provincial, prefectural, and county-level administrative divisions. The data includes: Provincial-level administrative regions: including 23 provinces, 5 autonomous regions, 4 municipalities directly under the Central Government, and 2 special administrative regions, a total of 34 provincial-level administrative regions. Prefectural-level administrative regions: including 293 prefecture-level cities, 7 regions, 30 autonomous prefectures, and 3 leagues, a total of 333 prefectural-level administrative regions. County-level administrative regions: including 977 city districts, 1301 counties, 394 county-level cities, 117 autonomous counties, 49 banners, 3 autonomous banners, 1 special zone, and 1 forest area, a total of 2843 county-level administrative regions; Classification by economic belt: By classifying each province into economic belts, including the eastern coastal area, the central inland area, and the western remote areas, a first-level classification is formed; Classification by administrative region: Using provinces for administrative region classification, including North China, Northeast China, East China, South China, Southwest China, and Northwest China, a second-level classification is formed; Classification by urban planning: Referring to the "Urban Planning Law of the People's Republic of China", a four-level classification is formed by classifying cities, including super-large cities, megacities, medium-sized cities, type-I small cities, type-I small towns, type-II cities, and type-II small towns; Classification by city type: Referring to the "City Commercial Charm Ranking List" (hereinafter referred to as the "list") released by the "New First-Tier City Research Institute" under Yicai on May 30, 2023. The list divides 337 cities across the country into the following five camps: first-tier cities (4), new first-tier cities (15), second-tier cities (30), third-tier cities (70), fourth-tier cities (90), and fifth-tier cities (128). Specifically, the commercial charm index is divided into five dimensions: agglomeration degree of commercial resources, urban hub nature, urban population activity, diversity of lifestyle, and future plasticity, and finally a five-level classification is formed; Rural / urban / town classification: It is mainly based on a seven-level classification and then distinguishes the address attributes, and the address types are distinguished as: urban area, city, urban village, rural area, banner township, banner town, district township, district town, district town village, county-level city township, county-level city town, county township, county village, county town, and county town village.

[0056] It can be explained that the detailed address classification is as follows: Natural Language Processing (NLP): The system uses NLP technology to analyze the text description in the address information. The NLP model will automatically identify the key elements in the address, such as street names, buildings, and locations. This step can help determine whether the address contains specific descriptive vocabulary to identify special location types, such as schools, government agencies, hotels, etc.; Rule Engine Application: The system utilizes an address rule engine, which applies a set of rules based on the elements in the address and the NLP results. For example, the rules include: If the address contains keywords such as "school" or "education", it is classified as a school-type address. If the address contains keywords such as "government" or "administration", it is classified as a government agency-type address. If the address contains keywords such as "hotel" or "guesthouse", it is classified as a hotel-type address. If the address contains keywords such as "express" or "logistics", it is classified as an express station-type address. If the address contains keywords such as "farm", "farmstead", or "agriculture", it is classified as a farm-type address. If the address contains keywords such as "village", "group", or "team", it is classified as a rural-type address; Determination of Address Detail Level: Based on the results of the rule engine, the system further determines the detail level of the address. For example, if the address contains keywords such as "community" or "neighborhood", it can be determined that the address is detailed to the community level. Then, after including the above detailed keywords, the system further determines whether there is a house number to distinguish whether it is detailed to the numbered level; Output of Classification Results: Finally, the system outputs the results of the address classification as different levels of address types and detail levels; Sub-classification of Address Details: Rule Engine Application: The address rule engine is applied again. The rule engine will apply a set of rules based on the keywords and context in the nine-level address to determine the classification to which the address belongs. For example: If the nine-level address contains keywords related to "school", it is classified as a school address. If the nine-level address contains keywords related to "farm", "forest farm", "fish farm", "ranch", it is classified as the corresponding type of farm / forest farm / fish farm / ranch address. If the nine-level address contains detailed descriptions, such as "the community has a detailed address", "the village has a number", etc., it is classified as a detailed address. If the nine-level address contains detailed descriptions, such as "warehouse", "express station", "hotel", etc., it is classified as a place address; Output of Classification Results: Finally, according to the results of the rule engine, the classification results of the nine-level address are output as the belonging classification, such as detailed address, place address, school address, farm / forest farm / fish farm / ranch address.

[0057] In this embodiment, the address to be classified is classified according to a preset classification strategy to obtain a first classified address; the first classified address is classified in detail according to a preset natural language processing method to obtain a second classified address; the address detail level corresponding to the second classified address is determined according to the second classified address; and the second classified address is classified in detail according to the address detail level to obtain a target classified address. Through hierarchical grouping from macro to micro, different address types are classified and graded, including classifications at multiple levels such as regional distribution, administrative regions, provinces, city sizes, city levels, city categories, regions, urban-rural attributes, and detailed address types, which improves the accuracy of correspondence address classification and reduces the cost of correspondence delivery.

[0058] Reference Figure 3 , Figure 3 is a schematic flowchart of the second embodiment of a correspondence address classification method of the present invention.

[0059] Based on the above first embodiment, the correspondence address classification method in this embodiment in step S20 includes:

[0060] Step S21: Identify address keywords from the first classified address according to preset keywords to obtain address keywords.

[0061] It can be understood that the preset keywords refer to street names, buildings, locations, etc., and the address keywords refer to the keywords included in the first classified address. The first classified address is identified according to the preset keyword types to obtain the keywords included in the first classified address.

[0062] Step S22: Determine the keyword type corresponding to the address keyword according to the address keyword.

[0063] It can be understood that the keyword type refers to the type of keywords in the first classified address.

[0064] Step S23: Classify the first classified address in detail according to the keyword type corresponding to the address keyword to obtain a second classified address.

[0065] It can be understood that the first classified address is classified in detail according to the type of keywords in the first classified address to obtain a second classified address.

[0066] It should be noted that, in order to accurately obtain the second classified address, further, the step of classifying the first classified address in detail according to the keyword type corresponding to the address keyword to obtain a second classified address includes: when the address keyword is a school, determining that the keyword type is a school type; marking the first classified address according to the school type to obtain an initial school address; and determining that the initial school address is the second classified address.

[0067] It is understandable that when the keyword in the first classification address is "school", the keyword type is determined as the school type, the first classification address is marked as the school type address according to the school type, and finally the second classification address is obtained according to the school type address.

[0068] It should be noted that in order to accurately obtain the address detail level, further, determining the address detail level corresponding to the second classification address according to the second classification address includes: determining a first detail level according to the address keyword corresponding to the second classification address; making a detailed level judgment on the second classification address according to the address keyword and the first detail level to obtain a second detail level; and determining the second detail level as the address detail level corresponding to the second classification address.

[0069] It is understandable that the first detail level refers to the current detail level of the second classification address, and the second detail level refers to the detail level of the second classification address after including the address keyword. For example: according to the result of the rule engine, the system further determines the detail level of the address. For example, if the address contains keywords such as "community" or "neighborhood", it can be determined that the address is detailed to the community level, and then after including the above detailed keywords, the system determines whether there is a house number to distinguish whether it is detailed to the level with a number.

[0070] It should be noted that in order to accurately obtain the first classification address, further, classifying the address to be classified according to the preset classification strategy to obtain the first classification address includes: obtaining administrative division information, economic belt division information, urban planning information, and township division information; classifying the address to be classified according to the administrative division information, the economic belt division information, the urban planning information, and the township division information to obtain the first classification address.

[0071] It is understandable that the administrative division information includes North China, Northeast China, East China, South China, Southwest China, and Northwest China; the economic belt division information includes the eastern coastal area, the central inland area, and the western remote area; the urban planning information includes super large cities, megacities, medium-sized cities, type-I cities, type-I small cities, type-II cities, and type-II small cities; and the township division information includes urban areas, cities, urban villages, rural areas, banner townships, banner towns, district townships, district towns, district town villages, county-level city townships, county-level city towns, county townships, county town villages, county towns, and county town villages.

[0072] It should be noted that as shown in Table 1, the address classification method of this embodiment includes a total of ten-level classification methods. The first-level classification (regional distribution: 3 levels): According to the province where the address is located, the addresses are divided into different regional distributions, such as the central inland region, the eastern coastal region, the western border region, etc.; the second-level classification (administrative region: 6 levels): Based on the division of administrative regions, the addresses are classified more precisely, such as Northeast, North China, East China, South China, Northwest, Southwest, etc.; the third-level classification (each province: 31 provinces): Further subdivide the addresses into each province to more accurately identify the geographical location of the addresses; the fourth-level classification (city scale: 8 levels): According to the city scale, the addresses are classified as super-large cities, megacities, medium-sized cities, type I cities, type I small cities, type II cities, type II small cities, others, etc., in order to better understand the scale and characteristics of the cities; the fifth-level classification (city level: 7 levels): The cities are divided into different levels, such as first-tier, new first-tier, second-tier, third-tier, fourth-tier, fifth-tier, other cities, in order to better understand the development stage and characteristics of the cities; the sixth-level classification (city category: 7 levels): The cities are classified as municipalities directly under the Central Government, provincial capital cities, prefecture-level cities, provincial-level administrative regions, autonomous prefectures, leagues, regions, etc., to identify the political status and organizational structure of the cities; the seventh-level classification (region: 17 levels): The addresses are divided into main urban areas, prefecture-level cities, county-level districts, counties, autonomous counties, ethnic counties, county-level cities, new areas, banners, autonomous banners, special zones, industrial parks, management areas, suburbs, development zones, mining areas, forest areas, to understand the location of the addresses within the city; the eighth-level classification (urban-rural: 15 levels): Distinguish whether the address is located in the urban area, city, urban village, rural area, banner township, banner town, district township, district town, district town and village, county-level city township, county-level city town, county township, county village, county town, county town and village, etc., to help understand the urban-rural attributes of the addresses; the ninth-level classification (detailed address type: 9 levels): Classify the addresses more specifically, such as place address, road area, road address, company address, other addresses, unable to judge, detailed address, school address, government agency address, etc., to better distinguish different types of addresses; the tenth-level classification (address detail level: 113 levels): Judge the detail level of the address, including: city - warehouse, express delivery; city - field; city - city; city - avenue without number; city - avenue with number; city - building; city - road / street + unknown address; city - store; city - team / group without number; city - team / group with number; city - highway; city - industrial area; city - industrial area with number or building; city - company; city - park / square without number; city - square with number; city - collective household; city - several districts and numbers; city - family courtyard / area / building; city - street / lane without number; city - street / lane with number; city - hotel, etc.; city - neighborhood committee / office; city - neighborhood committee / office + others; city - lane / block without number; city - lane / block with number; city - road without number; city - road with number; city - intersection with number; city - gate; city - others; city - others with number; city - district; city - market / mall;City—Depression / Pond etc. without number; City—Depression / Pond etc. with number; City—County / Hillock etc. without number; City—County / Hillock etc. with number; City—Residential area without detailed address; City—Residential area with detailed address; City—School; City—Garden; City—Courtyard; City—Government; Village + Other + number; Village + Other; Village without number; Village - Village Committee; Village - Village Committee + number; Village - Avenue without number; Village - Avenue with number; Village with number; Village - Team / Group without number; Village - Team / Group with number; Village - Industrial Zone; Village - Industrial Zone with number or building; Village - Street / Lane without number; Village - Street / Lane with number; Village - Neighborhood etc. without number; Village - Neighborhood etc. with number; Village - Road without number; Village - Road with number; Village - Road intersection without number; Village - Road intersection with number; Village - Depression / Pond etc. without number; Village - Depression / Pond etc. with number; Village - County etc. without number; Village - County etc. with number; Township - Without number; Township - Warehouse, Courier; Township - Field; Township - Avenue without number; Township - Avenue with number; Township - Building; Township - With number; Township - Road / Street + Unknown address; Township - Store; Township - Team / Group without number; Township - Team / Group with number; Township - Highway; Township - Industrial Zone with number or building; Township - Company; Township - Park / Square without number; Township - Square with number; Township - Collective household; Township - A certain district, a certain number; Township - Family courtyard / area / building; Township - Street / Lane without number; Township - Street / Lane with number; Township - Hotel etc.; Township - Neighborhood Committee / Office; Township - Neighborhood Committee / Office + Other; Township - Neighborhood etc. without number; Township - Neighborhood etc. with number; Township - Road without number; Township - Road with number; Township - Road intersection with number; Township - Gate; Township - Other; Township - Other with number; Township - District; Township - Market / Mall; Township - Depression / Pond etc. without number; Township - Depression / Pond etc. with number; Township - County / Hillock etc. without number; Township - County / Hillock etc. with number; Township - Residential area without detailed address; Township - Residential area with detailed address; Township - School; Township - Garden; Township - Courtyard; Township - Government.;

[0073] Table 1:

[0074]

[0075] In this embodiment, keyword recognition is performed on the first classified address according to a preset keyword to obtain an address keyword; the keyword type corresponding to the address keyword is determined according to the address keyword; and the first classified address is classified in detail according to the keyword type corresponding to the address keyword to obtain a second classified address. By recognizing the keywords in the first classified address to obtain an address keyword, then obtaining the keyword type of the address keyword, and classifying the first classified address in detail according to the keyword type to obtain a second classified address, the accuracy of the correspondence address classification is improved, and the delivery rate of the correspondence address is improved.

[0076] It should be understood that the above is only for illustration and does not constitute any limitation to the technical solution of the present invention. In specific applications, those skilled in the art can set according to needs, and the present invention does not make any restrictions on this.

[0077] It should be understood that although each step in the flowchart in the embodiments of the present application is shown sequentially according to the indication of the arrow, these steps are not necessarily executed sequentially according to the order indicated by the arrow. Unless there is a clear indication in this article, the execution of these steps has no strict order restriction, and they can be executed in other orders. Moreover, at least a part of the steps in the figure may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be executed alternately or alternately with at least a part of other steps or sub-steps or stages of other steps.

[0078] It should be noted that the above-described work process is only illustrative and does not constitute a limitation to the protection scope of the present invention. In actual applications, those skilled in the art can select some or all of them according to actual needs to achieve the purpose of the solution of this embodiment, and no restrictions are made here.

[0079] In addition, it should be noted that in this article, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or system including a series of elements not only includes those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such a process, method, article or system. Without further limitation, an element defined by the statement "including one..." does not exclude the existence of another identical element in the process, method, article or system including that element.

[0080] The serial numbers of the embodiments of the present invention above are only for description and do not represent the advantages or disadvantages of the embodiments.

[0081] The above is only the preferred embodiment of the present invention, and does not limit the patent scope of the present invention accordingly. Any equivalent structure or equivalent process transformation made by using the specification and drawings of the present invention, or directly or indirectly applied in other related technical fields, shall be included in the patent protection scope of the present invention by the same token.

Claims

1. A method for classifying correspondence addresses, characterized in that, The above-mentioned correspondence address classification method includes: Classify the address to be classified according to a preset classification strategy to obtain a first classified address; Perform a detailed address classification on the first classified address according to a preset natural language processing method to obtain a second classified address; Determine the address detail level corresponding to the second classified address according to the second classified address; Perform a detailed address classification on the second classified address according to the address detail level to obtain a target classified address.

2. The method according to claim 1, wherein The performing a detailed classification on the first classified address according to a preset natural language processing method to obtain a second classified address includes: Perform keyword recognition on the first classified address according to preset keywords to obtain address keywords; Determine the keyword type corresponding to the address keywords according to the address keywords; Perform a detailed classification on the first classified address according to the keyword type corresponding to the address keywords to obtain a second classified address.

3. The method according to claim 2, wherein The performing a detailed classification on the first classified address according to the keyword type corresponding to the address keywords to obtain a second classified address includes: When the address keyword is a school, determine that the keyword type is a school type; Mark the first classified address according to the school type to obtain an initial school address; Determine that the initial school address is the second classified address.

4. The method according to claim 1, wherein The determining the address detail level corresponding to the second classified address according to the second classified address includes: Determine a first detail level according to the address keyword corresponding to the second classified address; Perform a detailed level judgment on the second classified address according to the address keyword and the first detail level to obtain a second detail level; Determine that the second detail level is the address detail level corresponding to the second classified address.

5. The method according to claim 1, characterized in that, The performing a detailed address classification on the second classified address according to the address detail level to obtain a target classified address includes: Determine the classification keyword of the second classified address according to a preset classification rule; Determine the classification address type of the second classified address according to the classification keyword of the second classified address; Mark the second classified address according to the classification address type to obtain a target classified address.

6. The method according to claim 5, wherein The marking the second classified address according to the classification address type to obtain a target classified address includes: When the classification keyword is a warehouse keyword, determine that the classification address type is a place type; Mark the second classified address according to the place type to obtain a place type address; Determine that the place type address is the target classified address.

7. The method according to any one of claims 1 to 6, characterized in that The classifying the address to be classified according to a preset classification strategy to obtain a first classified address includes: Obtain administrative division information, economic belt division information, urban planning information, and township division information; Classify the address to be classified according to the administrative division information, the economic belt division information, the urban planning information, and the township division information to obtain a first classified address.

Images