Address matching method and device and computer storage medium

Abstract

The invention discloses an address matching method and device and a computer storage medium, and the method comprises the steps: 10, carrying out the analysis of a to-be-matched address through employing a natural language processing technology, and obtaining a standardized address which is arranged according to an address level; and step 20, carrying out step-by-step matching on the standardized address and the standard address base according to the address hierarchy to obtain a full-address matching result. According to the address matching method and device and the computer storage medium provided by the invention, the addresses in various formats can be matched, and the address matching speed and precision in a big data environment are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of address matching, and in particular to an address matching method and device and computer storage medium. BACKGROUND

[0002] In the field of data processing, especially in the fields of geographic information management, logistics distribution, customer relationship management system (CRM), etc., address matching is a key task. However, due to the diversity of address writing habits, even the description of the same place can appear in different forms, which poses a challenge to address matching. Traditional address matching methods rely on full character comparison, which is time-consuming and error-prone, especially when facing large-scale data sets, the performance bottleneck of this method is particularly obvious. SUMMARY

[0003] To solve the above technical problems, the present application provides an address matching method, device and computer storage medium, which can match addresses in multiple formats, improving the speed and accuracy of address matching in a big data environment.

[0004] The present application adopts the following technical solutions: In a first aspect, the present application provides an address matching method, comprising: Step 10, using natural language processing technology to analyze the address to be matched to obtain a standardized address arranged according to address levels; Step 20, according to the address levels, performing a step-by-step matching of the standardized address with a standard address library to obtain a full address matching result.

[0005] In combination with the first aspect, in a first possible implementation manner of the first aspect, the step 10 specifically comprises: Step 101, cleaning the address to be matched to remove redundant characters and spaces; Step 102, according to a self-defined address vocabulary table, screening out street names, house numbers, cities, provinces and countries from the cleaned address to be matched; or using a pre-trained language model to screen out street names, house numbers, cities, provinces and countries from the cleaned address to be matched; Step 103, arranging the screened street names, house numbers, cities, provinces and countries according to address levels to obtain a standardized address.

[0006] In combination with the first aspect, in a second possible implementation manner of the first aspect, the step 20 specifically comprises: Step 201, matching the address at the highest address level in the standardized address with the standard address library; Step 202, if the matching is successful, obtaining a sub-address matching result of the current address level, matching the next address level in the standardized address with the standard address library; if the matching is not successful, using a fuzzy matching method to match the address of the current address level, obtaining a sub-address matching result of the current address level, and matching the next address level in the standardized address with the standard address library; Step 203, performing step 202 until a sub-address matching result of the lowest address level is obtained, so as to obtain a full address matching result.

[0007] With reference to the second possible implementation manner of the first aspect, in a third possible implementation manner of the first aspect, the fuzzy matching method specifically includes: obtaining all standard addresses of the next address level of the sub-address matching result of the previous address level in the standard address library; calculating the similarity between the address of the current address level in the standardized address and all the standard addresses; selecting the standard address with the highest similarity as the sub-address matching result of the current address level.

[0008] With reference to the first aspect, in a fourth possible implementation manner of the first aspect, further including: Step 30, using other address data sources to perform secondary verification on the full address matching result; if the verification is successful, taking the full address matching result as the matching address; otherwise, re-performing the matching.

[0009] The second aspect of the present application provides an address matching device, including: a format unification module, configured to parse the address to be matched by using a natural language processing technology, and obtain a standardized address arranged according to address levels; a step-by-step matching module, configured to perform step-by-step matching between the standardized address and a standard address library according to the address levels, and obtain a full address matching result.

[0010] With reference to the second aspect, in a first possible implementation manner of the second aspect, the format unification module includes: a preprocessing submodule, configured to clean the address to be matched, and remove redundant characters and spaces; a screening submodule, configured to screen out street names, house numbers, cities, provinces and countries from the cleaned address to be matched according to a self-defined address vocabulary table; or use a pre-trained language model to screen out street names, house numbers, cities, provinces and countries from the cleaned address to be matched; an arrangement submodule, configured to arrange the screened street names, house numbers, cities, provinces and countries according to address levels, and obtain a standardized address.

[0011] With reference to the second aspect, in a second possible implementation form of the second aspect, the step-by-step matching module comprises a fuzzy matching submodule: obtaining all standard addresses of a next address level of a sub-address matching result of a previous address level in the standard address library; calculating the similarity between the address of the current address level in the standardized address to be matched and all standard addresses; and selecting the standard address with the highest similarity as the sub-address matching result of the current address level.

[0012] With reference to the second aspect, in a third possible implementation form of the second aspect, the method further comprises: a verification module configured to quote a reliable address data source to perform secondary verification on the full address matching result.

[0013] In a third aspect, the present application provides a computer storage medium, which stores a computer program, and the computer program is executed to perform the address matching method provided in the first aspect.

[0014] The address matching method, device and computer storage medium provided by the present application can match addresses of various formats, and improve the speed and accuracy of address matching in a big data environment. Compared with the prior art, in the present application, the natural language processing technology is used to analyze the address to be matched first to obtain the standardized address arranged according to the address level; then the standardized address is matched with the standard address library according to the address level to obtain the full address matching result. By analyzing addresses of various formats to obtain standardized addresses of a unified format for matching, the matching difficulty is simplified, and various format addresses can be matched; the step-by-step matching according to the address level can avoid unnecessary full character comparison, greatly reduce the calculation amount, and improve the speed and accuracy of address matching. BRIEF DESCRIPTION OF DRAWINGS

[0015] In order to more clearly illustrate the technical solutions of the present application, the following will briefly introduce the drawings needed to be used in the present application. Obviously, the drawings described below are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0016] Figure 1 The flow chart of the address matching method provided by the embodiment of the present application; Figure 2 The structural block diagram of the address matching device provided by the embodiment of the present application. DETAILED DESCRIPTION

[0017] In order to make the technical solution of the present application better understood by those skilled in the art, the present application will be described in further detail below in combination with the drawings and specific embodiments. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, the embodiments obtained by those skilled in the art without making creative efforts in advance should belong to the protection scope of the present application.

[0018] In the existing address matching method, full character comparison is mostly used, and the full characters of the address to be matched are compared with the address library, and whether the characters of the two address strings are completely consistent is directly compared. However, when the size of the address library is large, the calculation amount of character-by-character comparison of all candidate addresses is large, and the efficiency is low. It is sensitive to spelling errors, abbreviations, and dialect expressions (such as "Zhejiang" vs. "Zhejiang"), which leads to missed matching, poor fault tolerance, and low matching accuracy.

[0019] In the embodiments of the present application, natural language processing technology is used to analyze the address to be matched first, and a standardized address arranged according to the address level is obtained; then the standardized address is matched with the standard address library level by level according to the address level, and a full address matching result is obtained. By analyzing a plurality of different formats of addresses, a standardized address in a unified format is obtained for matching, which simplifies the matching difficulty and can realize matching of a plurality of formats of addresses; the level-by-level matching according to the address level can avoid unnecessary full character comparison, greatly reduce the calculation amount, and improve the speed and accuracy of address matching.

[0020] The method flow provided in the embodiments of the present application can be specifically run in a server. The server includes input unit, processor unit, output unit, communication unit, storage unit, peripheral unit and the like components. These components communicate through one or more buses. In the present application, the server can be a server for managing address information of a geographic information management center, a logistics distribution center or a customer relationship management center, or can be a workstation device used by technical personnel. Specifically: The input unit is used to realize the interaction between the user and the server and / or the input of information into the server. For example, the input unit can receive the digital or character information input by the user to generate a signal input related to user settings or function control. In the specific embodiments of the present application, the input unit can be a touch panel, or other human-computer interaction interfaces such as physical input keys, a mouse or a joystick.

[0021] The processor unit is the control center of the server, connects various parts of the server by various interfaces and lines, executes various functions of the server and / or processes data by running or executing software programs and / or modules stored in the storage unit and calling data stored in the storage unit. The processor unit can be composed of an integrated circuit (IC), for example, it can be composed of a single packaged IC, or it can be composed of multiple packaged ICs with the same function or different functions connected. For example, the processor unit can only include a central processing unit (CPU), or it can be a combination of a GPU, a digital signal processor (DSP), and a control chip in a communication unit (such as a baseband chip). In the embodiments of the present application, the CPU can be a single operation core or can include multiple operation cores.

[0022] The communication unit is used to establish a communication channel, so that the server can be connected to a remote device through the communication channel and interact with the remote device. The remote device is, for example, a terminal device of a staff, which can send a request message to the server to trigger the server to start executing the method flow provided in the embodiment.

[0023] The output unit includes but is not limited to an image output unit and a sound output unit. The image output unit is used to output text, pictures and / or videos. The image output unit can include a display panel, for example, a display panel configured in the form of an LCD (Liquid Crystal Display), an OLED (Organic Light-Emitting Diode), a field emission display (FED), etc.

[0024] The storage unit can be used to store software programs and modules, and the processing unit executes various functions of the server and implements data processing by running the software programs and modules stored in the storage unit. The storage unit mainly includes a program storage area and a data storage area, wherein the program storage area can store an operating system and at least one application program required by a function, such as an address matching program. In the specific embodiments of the present application, the storage unit can include a volatile memory, such as a non-volatile random access memory (NVRAM), a phase change random access memory (PRAM), a magetoresistive random access memory (MRAM), and the like, and can also include a non-volatile memory, such as at least one disk storage device, an electrically erasable programmable read-only memory (EEPROM), a flash memory device, such as a NOR flash memory or a NAND flash memory. The non-volatile memory stores the operating system and the application program executed by the processing unit. The processing unit loads the running program and data from the non-volatile memory to the memory and stores the digital content in the mass storage device.

[0025] The power supply is used to supply power to different components of the server to maintain its operation, including an external power supply directly supplying power to the server, such as an AC adapter and the like. In some embodiments of the present application, the power supply can also be defined more widely, for example, it can also include a power management system, a charging system, a power failure detection circuit, a power converter or inverter, a power state indicator (such as a light-emitting diode), and any other components associated with the generation, management and distribution of electrical energy of the server.

[0026] The embodiment of the present application provides an address matching method, as shown in the following schematic diagram, which comprises the following steps: Figure 1 Step 10: using a natural language processing technology to parse the address to be matched to obtain a standardized address arranged according to an address level; Step 20: performing a step-by-step matching between the standardized address and a standard address library according to the address level to obtain a full address matching result.

[0027] Preferably, step 10 specifically comprises the following steps: Step 101: cleaning the address to be matched to remove redundant characters and spaces.

[0028] ​Step 102, according to the self-defined address vocabulary, the street name, house number, city, province, country are screened from the cleaned address to be matched. The address vocabulary includes country name, province name, city name, district name, street name, etc.

[0029] Alternatively, using a pre-trained language model, the house number, street name, district, city, province, country are screened from the cleaned address to be matched.

[0030] Step 103, arrange the screened house number, street name, district, city, province, country according to the address level to obtain the standardized address. Preferably, arrange according to the address level from high to low, that is, national level, provincial level, municipal level, district level, street level, house number level.

[0031] For example, the address to be matched is "Beijing Chaoyang District Jianguo Road No. 1", and the standardized address obtained after step 10 is "China | Beijing | Chaoyang District | Jianguo Road | No. 1". The address to be matched is "Chongqing Chongqing Qianjiang District Chongqing-Qianjiang District Jia Yi Street Bing Ding West Road No. 100", and the standardized address obtained after step 10 is "China | Chongqing | Qianjiang District | Jia Yi Street | Bing Ding West Road | No. 100". The address to be matched is "Sichuan Province Chengdu Shuangliu District Chengdu-Tianfu New District Jia Yi Street Jia Yi Avenue South Section No. 110", and the standardized address obtained after step 10 is "China | Sichuan | Chengdu | Shuangliu District, Tianfu New District | Jia Yi Avenue | No. 110".

[0032] The embodiment of the application first analyzes the address to be matched, screens the house number, street name, district, city, province, and country, and arranges them according to the address level to obtain the standardized address. The address to be matched in different formats is unified to the same format and arranged by level, which is convenient for matching.

[0033] Preferably, step 20 specifically includes: Step 201, match the highest address level in the standardized address with the standard address library. The standard address library is constructed by collecting official address data published by government agencies, postal code database or address data in geographic information system.

[0034] Step 202, if the matching is successful, the sub-address matching result of the current address level is obtained, and the next address level in the standardized address is matched with the standard address library. If the matching is not successful, the fuzzy matching method is used to match the address of the current address level to obtain the sub-address matching result of the current address level, and the address of the next address level in the standardized address is matched with the standard address library.

[0035] Step 203, step 202 is executed until the lowest address level sub-address matching result is obtained, so as to obtain the full address matching result.

[0036] For example, the standardized address is "China | Chongqing | Qianjiang District | Jia Yi Street | Bing Ding West Road | No. 100", the country level "China" is matched first, then the province level "Chongqing", and so on, until the matching process of the whole standardized address is completed, and the matching result is "China | Chongqing | Qianjiang District | Jia Yi Street | Bing Ding West Road | No. 100". The standardized address is "China | Sichuan | Chengdu | (Shuangliu District, Tianfu New Area) | Jia Yi Avenue | No. 110", the country level "China" is matched first, then the province level "Sichuan", and then the city level "Chengdu". When the district and county level is matched, since the same level address appears, the last appearing address "Tianfu New Area" is selected. In this way, until the matching process of the whole standardized address is completed, the full address matching result is "China | Sichuan | Chengdu | Tianfu New Area | Jia Yi Avenue | No. 110".

[0037] Preferably, in step 202, the fuzzy matching method specifically comprises the following steps: All standard addresses of the next address level of the sub-address matching result of the previous address level in the standard address library are obtained. The similarity of the address of the current address level in the standardized address and all standard addresses is calculated. The standard address with the highest similarity is selected as the sub-address matching result of the current address level.

[0038] For example, the standardized address is "China | Chongqing | Economic Development Zone | Changsheng Bridge Town | Tea Flower Road | No. 1", when the district and county level "Economic Development Zone" is matched, the matching with the district and county level standard address "Economic Development Zone" in the standard address library is unsuccessful. At this time, all district and county level standard addresses under Chongqing city in the standard address library are obtained, for example, "Economic Development Zone", "Yuzhong District", "Wanzhou District", "Fuling District", "Dadukou District", "Jiangbei District", "Shapingba District", etc. The similarity of the district and county level address "Economic Development Zone" in the standardized address to be matched and each district and county level standard address is calculated, that is, the percentage of the number of same characters in the total number of characters is calculated. Finally, all similarities are compared, and the district and county level standard address "Economic Development Zone" with the highest similarity is selected as the matching address of the district and county level address "Economic Development Zone" in the standardized address to be matched.

[0039] In the embodiment of the application, the standardized address in the unified format is matched level by level from high to low according to the address level, without full character comparison of the whole address, so that the calculation amount is greatly reduced, and the speed and accuracy of address matching are improved. When each address level is matched, the address standard library is matched first, if the matching is unsuccessful, the fuzzy method is used for matching, so that the situation that direct matching is unsuccessful due to the use of abbreviations, acronyms and other formats by part of the address can be avoided, and the matching accuracy is greatly improved.

[0040] Preferably, the address matching method provided by the embodiment of the present application further comprises: In step 30, the full address matching result is verified by using other address data sources. If the verification is successful, the full address matching result is taken as the matching address; otherwise, the matching is performed again.

[0041] The other address data sources are different from the data sources for constructing the standard address library, and can be data interfaces of a map platform open source.

[0042] The embodiment cross-uses other reliable address data sources to verify the matching structure, thereby further ensuring the accuracy of the matching and improving the matching precision.

[0043] A specific example is provided below.

[0044] The address to be matched is "No. 110, Jiayi Road, Cihu High-tech Development Zone, Huashan District, Ma'anshan City, Anhui Province, China".

[0045] The address to be matched is cleaned to remove redundant characters and spaces. According to a self-defined address vocabulary table, the street name "Jiayi Road", the house number "No. 110", the district and county "Cihu High-tech Development Zone" and "Huashan District", the city "Ma'anshan", the province "Anhui", and the country "China" are filtered from the cleaned address to be matched. The filtered house number, street name, district and county, city, province, and country are arranged according to the address hierarchy to obtain the standardized address "China | Anhui | Ma'anshan | Huashan District, Cihu High-tech Development Zone | Jiayi Road | No. 110".

[0046] The country "China" is matched first, then the province "Anhui", and then the city "Ma'anshan". When the district and county level is matched, since the same level address appears, the last appearing address "Cihu High-tech Development Zone" is selected. However, the "Cihu High-tech Development Zone" does not match the district and county level standard address "Cihu High-tech Industrial Development Zone" in the standard address library. At this time, all district and county level standard addresses under Ma'anshan City in the standard address library are obtained, and the similarity of the district and county level address "Cihu High-tech Development Zone" in the standardized address to be matched with each district and county level standard address is calculated, i.e., the percentage of the number of identical characters to the total number of characters. Finally, all similarities are compared, and the district and county level standard address "Cihu High-tech Industrial Development Zone" with the highest similarity is selected as the matching address of the district and county level address "Cihu High-tech Development Zone" in the standardized address to be matched. The matching is performed in turn until the entire standardized address matching process is completed, and the matching result is "China | Anhui | Ma'anshan | Cihu High-tech Industrial Development Zone | Jiayi Road | No. 110".

[0047] The address matching method is checked again by using the Baidu address library platform, and if the checking is successful, the matched address is "China | Anhui | Ma'anshan | Cihu High-tech Industrial Development Zone | Jiayi Road | No.

[0048] The address matching method provided by the embodiment of the application simplifies the matching difficulty, can realize matching of addresses in multiple formats, can avoid unnecessary full character comparison, greatly reduces the calculation amount, and improves the speed and accuracy of address matching.

[0049] The embodiment of the application further provides an address matching device, as shown in Figure 2 The address matching device comprises: A format unification module is configured to parse the address to be matched by using a natural language processing technology to obtain standardized addresses arranged according to address levels.

[0050] A step-by-step matching module is configured to perform step-by-step matching of the standardized addresses and a standard address library according to the address levels to obtain a full address matching result.

[0051] The address matching device provided by the embodiment of the application can simplify the matching difficulty, can realize matching of addresses in multiple formats, can avoid unnecessary full character comparison, greatly reduces the calculation amount, and improves the speed and accuracy of address matching.

[0052] Preferably, the format unification module comprises: A preprocessing submodule is configured to clean the address to be matched to remove redundant characters and spaces.

[0053] A screening submodule is configured to screen a street name, a house number, a city, a province and a country from the cleaned address to be matched according to a self-defined address vocabulary table, or to screen a street name, a house number, a city, a province and a country from the cleaned address to be matched by using a pre-trained language model.

[0054] An arrangement submodule is configured to arrange the screened street name, house number, city, province and country according to the address levels to obtain the standardized addresses.

[0055] Preferably, the step-by-step matching module comprises a fuzzy matching sub-module: obtaining all standard addresses of the next address level of the sub-address matching result of the previous address level in the standard address library; calculating the similarity of the address of the current address level in the standardized address to be matched and all standard addresses; selecting the standard address with the highest similarity as the sub-address matching result of the current address level.

[0056] Preferably, the address matching device of the embodiment of the present application further comprises: The verification module is configured to quote a reliable address data source to perform secondary verification on the full address matching result.

[0057] The embodiment of the present application also provides a computer storage medium, which stores a computer program, and the computer program is executed to perform the address matching method provided by the embodiment of the present application.

[0058] The computer storage medium provided by the embodiment of the present application stores a computer program, and the computer program is executed to perform the address matching method. The address is processed to obtain a standardized address in a unified format for matching, so that the matching difficulty is simplified, and multiple format addresses can be matched; step-by-step matching is performed according to the address level, unnecessary full character comparison can be avoided, the calculation amount is greatly reduced, and the speed and accuracy of address matching are improved.

[0059] Each of the embodiments in the specification is described in a progressive manner, and the same and similar parts of each embodiment can be referred to. Each embodiment focuses on the difference from other embodiments. Especially, the device embodiment is basically similar to the method embodiment, so the description is relatively simple, and the related parts can be referred to the part of the method embodiment. Those skilled in the art can understand that the modules in the device can be adaptively changed and arranged in one or more devices different from the embodiment. The modules or units or components in the embodiment can be combined into one module or unit or component, and furthermore, they can be divided into multiple sub-modules or sub-units or sub-components. All features disclosed in the specification (including the accompanying claims, abstract and drawings) and all processes or units of any method or device disclosed in the specification can be combined in any combination, except that at least some of the features and / or processes or units are mutually exclusive. Unless explicitly stated otherwise, each feature disclosed in the specification (including the accompanying claims, abstract and drawings) can be replaced by an alternative feature providing the same, equivalent or similar purpose.

[0060] The above merely illustrates the specific embodiments of the present application, but the protection scope of the present application is not limited thereto, any person skilled in the art can easily think of the changes or replacements within the technical range disclosed by the present application, which should be covered in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. An address matching method, characterized in that, include: Step 10: Use natural language processing technology to parse the address to be matched to obtain a standardized address arranged in address hierarchy; Step 20: Match the standardized address with the standard address database level by level according to the address hierarchy to obtain the full address matching result.

2. The address matching method according to claim 1, characterized in that, Step 10 specifically includes: Step 101: Clean the address to be matched by removing extra characters and spaces; Step 102: Based on a custom address vocabulary, filter out street names, house numbers, cities, provinces, and countries from the cleaned addresses to be matched; or, use a pre-trained language model to filter out street names, house numbers, cities, provinces, and countries from the cleaned addresses to be matched. Step 103: Arrange the selected street names, house numbers, cities, provinces, and countries according to address hierarchy to obtain standardized addresses.

3. The address matching method according to claim 1, characterized in that, Step 20 specifically includes: Step 201: Match the highest address level in the standardized address set with the standard address database; Step 202: If the match is successful, obtain the sub-address matching result of the current address level, and match the address of the next address level in the standardized address with the standard address library; if the match is unsuccessful, use the fuzzy matching method to match the address of the current address level, obtain the sub-address matching result of the current address level, and match the address of the next address level in the standardized address with the standard address library. Step 203: Execute step 202 until the sub-address matching result of the lowest address level is obtained, thereby obtaining the full address matching result.

4. The address matching method according to claim 3, characterized in that, The fuzzy matching method specifically includes: Retrieves all standard addresses in the next address level of the sub-address matching result of the previous address level in the standard address library; Calculate the similarity between the address at the current address level in the normalized address set and all normalized addresses; The standard address with the highest similarity is selected as the sub-address matching result for the current address level.

5. The address matching method according to claim 1, characterized in that, Also includes: Step 30: Use other address data sources to perform a second verification on the full address matching results; If the verification is successful, the full address match result will be used as the matching address; Otherwise, perform the matching again.

6. An address matching device, characterized in that, include: The format unification module is used to parse the address to be matched using natural language processing technology to obtain standardized addresses arranged in address hierarchy; The step-by-step matching module is used to match the standardized address with the standard address database step by step according to the address hierarchy to obtain the full address matching result.

7. The address matching device according to claim 6, characterized in that, The format unification module includes: The preprocessing submodule is used to clean the addresses to be matched, removing extra characters and spaces; The filtering submodule is used to filter street names, house numbers, cities, provinces, and countries from the cleaned addresses to be matched based on a custom address vocabulary; or, using a pre-trained language model, to filter street names, house numbers, cities, provinces, and countries from the cleaned addresses to be matched. The sorting submodule is used to sort the filtered street names, house numbers, cities, provinces, and countries according to address hierarchy to obtain standardized addresses.

8. An address matching device according to claim 6, characterized in that, The step-by-step matching module includes a fuzzy matching submodule: used to obtain all standard addresses of the next address level of the sub-address matching result of the previous address level in the standard address library; calculate the similarity between the address of the current address level in the standardized address to be matched and all standard addresses; and select the standard address with the highest similarity as the sub-address matching result of the current address level.

9. An address matching device according to claim 6, characterized in that, Also includes: The verification module is used to reference a reliable address data source and perform secondary verification on the full address matching results.

10. A computer storage medium storing a computer program thereon, characterized in that, When the computer program is executed, it performs the address matching method according to any one of claims 1 to 5.

Images