Address processing method, device, medium and product

Abstract

The invention provides an address processing method and device, a medium and a product, and the method comprises the steps: obtaining an address text provided by an address provider in response to an address processing demand of the address provider; calling a pre-established model library, selecting an entity extraction model matched with the region information recorded by the address text from the model library, and inputting the address text into the entity extraction model to obtain an address entity result; the address entity result comprises address entities corresponding to different levels of addresses, and the address granularities of the different levels of addresses are different; calling a pre-established full-amount address base, and retrieving longitude and latitude information matched with the address entity result based on the full-amount address base; and feeding back the retrieved longitude and latitude information to the address provider.

Description

Technical Field

[0001] This specification relates to one or more embodiments in the field of logistics technology, and more particularly to an address processing method, apparatus, medium, and product. Background Technology

[0002] In logistics, particularly in pickup and last-mile delivery scenarios, it's crucial to match latitude and longitude coordinates based on the address information provided by merchants or consumers to prevent delivery personnel from being unable to pick up or deliver packages due to incorrect address information. However, in countries with poor logistics infrastructure, the lack of basic address infrastructure, significant differences in address information input by merchants and consumers, and the use of multiple languages ​​for address information make it difficult to match accurate latitude and longitude coordinates based on the provided address. This hinders delivery personnel from successfully completing pickup or delivery tasks, impacting logistics efficiency and user experience. Summary of the Invention

[0003] In view of this, in order to at least solve the technical problems of low logistics efficiency and poor user experience caused by the poor quality of logistics infrastructure services in some countries, this specification provides one or more embodiments of the following technical solutions:

[0004] According to a first aspect of one or more embodiments of this specification, an address processing method is provided, comprising:

[0005] In response to the address processing request from the address provider, obtain the address text provided by the address provider;

[0006] A pre-established model library is invoked, and an entity extraction model that matches the regional information recorded in the address text is selected from the model library. The address text is then input into the entity extraction model to obtain the address entity results. The address entity results include address entities corresponding to different levels of addresses, and the address granularity of different levels of addresses is different.

[0007] Call the pre-established full address database and retrieve latitude and longitude information that matches the address entity result based on the full address database;

[0008] The retrieved latitude and longitude information is fed back to the address provider.

[0009] According to a second aspect of one or more embodiments of this specification, an address processing apparatus is provided, comprising:

[0010] The acquisition module is configured to: in response to the address processing request of the address provider, acquire the address text provided by the address provider;

[0011] The extraction module is configured to: call a pre-established model library, select an entity extraction model from the model library that matches the regional information recorded in the address text, and input the address text into the entity extraction model to obtain address entity results; the address entity results include address entities corresponding to different levels of addresses, and the address granularity of different levels of addresses is different;

[0012] The retrieval module is configured to: call a pre-established full address database and retrieve latitude and longitude information that matches the address entity result based on the full address database;

[0013] The feedback module is configured to provide the retrieved latitude and longitude information to the address provider.

[0014] According to a third aspect of one or more embodiments of this specification, an electronic device is provided, comprising: a processor; a memory for storing processor-executable instructions; wherein the processor performs the executable instructions to implement the steps of the address processing method provided in the first aspect above.

[0015] According to a fourth aspect of one or more embodiments of this specification, a computer-readable storage medium is provided that stores computer instructions thereon, which, when executed by a processor, implement the steps of the address processing method provided in the first aspect above.

[0016] According to a fifth aspect of one or more embodiments of this specification, a computer program product is provided, comprising a computer program that, when executed by a processor, implements the steps of the address processing method provided in the first aspect above.

[0017] As can be seen from the above embodiments, this specification uses an entity extraction model that matches the regional information recorded in the address text provided by the address provider to identify the address text and divide it into different levels, resulting in address entity results containing address entities with different address granularities. This achieves better standardization of address texts with varying expressions into address entity results by utilizing named entity recognition and model combination. Furthermore, since the full address database is pre-established based on the mapping relationship between address entities and their latitude and longitude information, accurate latitude and longitude information can be retrieved and fed back to the address provider based on the address entity results. This allows couriers to complete tasks smoothly based on accurate latitude and longitude information during pickup or delivery, without repeatedly asking customers for addresses or engaging in multiple trial and error attempts, thus improving logistics efficiency and user experience. In addition, since the address entity results contain address entities at different address levels, the process of retrieving latitude and longitude information based on the full address database can also be performed layer by layer according to address levels, which can improve retrieval efficiency to a certain extent. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the architecture of an address processing service system provided in an exemplary embodiment.

[0019] Figure 2 This is a flowchart of an address processing method provided in an exemplary embodiment.

[0020] Figure 3 This is a flowchart of another address processing method provided in an exemplary embodiment.

[0021] Figure 4 This is a schematic diagram of the structure of a device provided in an exemplary embodiment.

[0022] Figure 5 This is a block diagram of an address processing apparatus provided in an exemplary embodiment. Detailed Implementation

[0023] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this specification, and not all embodiments. Based on the embodiments in this specification, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this specification.

[0024] The user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this manual are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, use and processing of related data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and corresponding operation portals are provided for users to choose to authorize or refuse.

[0025] To address the technical problems of low logistics efficiency and poor user experience caused by poor logistics infrastructure in some countries, this specification provides one or more embodiments of an address processing method. This method involves calling an entity extraction model that matches the regional information recorded in the address text provided by the address provider. The address text is then identified and divided into different levels, resulting in address entity results containing address entities with varying address granularities. This utilizes named entity recognition and model combination to better standardize address texts with diverse expressions into address entity results at different address levels. Furthermore, since the full address database is pre-established based on the mapping relationship between address entities and their latitude and longitude information, accurate latitude and longitude information can be retrieved from the full address database based on the address entity results and fed back to the address provider. This allows couriers to complete pickup or delivery tasks smoothly based on accurate latitude and longitude information, eliminating the need for repeated inquiries about customer addresses or multiple trial-and-error processes, thus improving logistics efficiency and user experience. Furthermore, since the address entity results contain address entities at different address levels, the process of retrieving latitude and longitude information based on the full address database can also be carried out by searching layer by layer according to the address level, which can improve the retrieval efficiency to a certain extent.

[0026] Figure 1 This is a schematic diagram of the architecture of an address processing service system provided in an exemplary embodiment. For example... Figure 1 As shown, the system may include a server 11, a network 12, and several electronic devices, such as a PC (Personal Computer) 13, a mobile phone 14, etc.

[0027] Server 11 can be a physical server containing an independent host, or it can be a virtual server hosted in a host cluster. During operation, server 11 can run server-side programs for a certain application to implement the relevant functions of that application. For example, when server 11 runs an address processing service program, it can function as a corresponding address processing service platform.

[0028] PC13 and mobile phone14 are just some of the types of electronic devices that users can use. In reality, users can obviously also use electronic devices such as tablets, laptops, PDAs (Personal Digital Assistants), wearable devices (such as smart glasses, smartwatches, etc.), etc., and one or more embodiments in this specification do not limit this. During operation, the electronic device can run a client-side program of an application to achieve the relevant functions of that application. For example, when the electronic device runs an address processing service program, it can act as a client for that address processing service. The client application for the address processing service can be launched and run on the electronic device. This client-side program can be a native application installed on the electronic device, or it can be a mini-program, quick app, or other similar form. Of course, when using web technologies such as HTML5 or similar, the relevant functions can be achieved through a page displayed by a browser. This browser can be a standalone browser application or a browser module embedded in some applications.

[0029] As for the network 12 that enables interaction between electronic devices such as PC13 and mobile phone 14 and server 11, communication can be achieved using either wired or wireless networks, depending on the communication methods supported by the respective electronic devices. This specification does not impose any restrictions on this. For example, PC13 can support both wired and wireless communication, so it can use either wired or wireless networks as needed. Mobile phone 14 typically only supports wireless communication, so it can use a wireless network for communication.

[0030] The address processing method provided in one or more embodiments of this specification can be applied to the server 11 described above. Please refer to [link / reference]. Figure 2 , Figure 2 This is a flowchart of an address processing method provided in an exemplary embodiment, the method including the following steps:

[0031] In step S100, in response to the address processing request from the address provider, the address text provided by the address provider is obtained;

[0032] In step S220, a pre-established model library is invoked, and an entity extraction model that matches the regional information recorded in the address text is selected from the model library. The address text is then input into the entity extraction model to obtain the address entity result. The address entity result includes address entities corresponding to different levels of addresses, and the address granularity of different levels of addresses is different.

[0033] In step S300, a pre-established full address database is invoked, and latitude and longitude information matching the address entity result is retrieved based on the full address database;

[0034] In step S410, the retrieved latitude and longitude information is fed back to the address provider.

[0035] When the sender submits a courier order, the logistics system receives the courier order and the pickup and delivery addresses recorded on it. It then sends an address processing request to the server 11, triggering the server 11 to execute steps S100 to S410. The logistics system in this context refers to the address provider.

[0036] During step S100, in response to the address processing request from the address provider, the address text provided by the address provider is obtained. After the address provider submits a logistics order, the logistics system can obtain the pickup address text and the recipient address text based on the sender and recipient addresses recorded in the logistics order. Therefore, during step S100, the address text provided by the address provider can be obtained by querying the logistics system. The address text can include both the pickup address text and the recipient address text. For ease of description, and because the processing logic for the pickup address text and the recipient address text is the same, both are referred to as address text.

[0037] Each address text obtained after executing step S100 can be processed through the above steps S220 to S410.

[0038] During step S220, a pre-established model library is invoked to select an entity extraction model that matches the regional information recorded in the address text. The address text is then input into the entity extraction model for processing, resulting in address entity results including address entities corresponding to different levels of addresses.

[0039] The model library can store multiple entity extraction models and the mapping relationship between each entity extraction model and regional information. Different entity extraction models can be trained on different training sets. In some examples, one entity extraction model can be applied to multiple regions or only one region. Based on this, the regional information in step S220 can be understood as the country information recorded in the address text, such as China, Russia, Vietnam, and Thailand, or it can be understood as the region to which the country belongs, such as Europe and the Americas or Southeast Asia. Since the address text corresponding to logistics orders records country information, the country information can be obtained based on the address text, and then the matching entity extraction model can be selected from the model library based on the country information.

[0040] In some embodiments, to improve the accuracy of the entity extraction results output by the entity extraction model and the generalization ability of the entity extraction model, the address processing method provided in one or more embodiments of this specification also provides a training set for the entity extraction model. That is, the entity extraction model is trained based on a pre-established address training set. The address training set includes multiple sets of training samples. Each set of training samples includes training address text and entity labels that match the training address text. The entity labels include address entities corresponding to the different address levels mentioned above.

[0041] The training address text can be obtained based on the address information recorded by the logistics system. For example, if the address information is "Country A, Province B, City C, District D, Road E, No. G", the address text can include the above address information. Or, if the address information is "Country B, City C, District D, Road E, No. G", the address text can include country information in addition to address information. The country information can be parsed by the logistics system based on the address information. The parsing principle can be found in relevant technologies, which will not be explained in detail here.

[0042] Taking an address text including "Country A, Province B, City C, District D, Road E, Number G" as an example, the different levels of address entities contained in the matching entity tags are: Country: A, Province: B, City: C, District: D, and Road: E. In some examples, to facilitate the storage of entity tags and simplify the model output, entity tags can be recorded in the form of a sequence, such as {Country: A, Province: B, City: C, District: D, Road: E}, but it is not limited to this.

[0043] It should be understood that the address hierarchy can be set according to actual needs. For example, it can be divided from the country level to the road level, or from the country level to the district level. This specification does not limit this. Accordingly, to improve retrieval efficiency and accuracy, the full address database can store latitude and longitude information layer by layer based on the above address hierarchy.

[0044] Therefore, by training the initial entity extraction model using the aforementioned address training set, the final trained entity extraction model can be better applied to the extraction of address entities in countries or regions that match the address training set, which is beneficial to improving the accuracy of the model output and the model's generalization ability.

[0045] In the process of training an initial entity extraction model based on an address training set from multiple countries or regions, if the address formats of the multiple countries or regions are inconsistent or irregular, it may lead to knowledge confusion in the initial entity extraction model during training, affecting the model's output accuracy and robustness, and thus hindering the accurate retrieval of latitude and longitude information based on address entities. Therefore, to solve this technical problem, in some embodiments, one or more embodiments of this specification also provide an optimization scheme for the address training set used by the initial entity extraction model. That is, an entity extraction model is trained based on an address training set whose address format similarity meets a set similarity threshold, wherein the similarity threshold can be set according to experience or actual needs. Understandably, for an entity extraction model, its corresponding address training set can be formed by addresses from countries with the same address format or the same pattern. In some examples, to better ensure the model's output accuracy and robustness, address training sets can be established by country, and then an initial entity extraction model can be independently configured for each country and trained independently to obtain the corresponding entity extraction model. That is, one country corresponds to one entity extraction model. This can reduce the interference of address training sets from different countries on the logical reasoning of the same entity extraction model, and better improve the model's output accuracy and robustness.

[0046] In some embodiments, to better improve the output accuracy, output efficiency and robustness of the entity extraction model, the entity extraction model can be a BERT (Bidirectional Encoder Representations from Transformers) model.

[0047] During the establishment of the address training set, a full address database can be established simultaneously. The address text and latitude / longitude information in the full address database can be mapped hierarchically. For example, the country can be stored in the parent node, the province in the next child node of the corresponding parent node, the city in the next child node of the corresponding province, and the district and its matching latitude / longitude in the next child node of the corresponding city. After storing the corresponding data step by step according to the address hierarchy, the mapping relationship between address entities and latitude / longitude information is obtained and stored in a database, thus establishing the full address database.

[0048] Based on this, during the execution of step S300, latitude and longitude information matching the address entity result can be retrieved layer by layer from the full address database. This layer-by-layer retrieval method does not require accessing all the data in the full address database, thus ensuring retrieval efficiency to a certain extent.

[0049] Then, step S410 is executed to send the retrieved latitude and longitude information back to the address provider. This allows the address provider to provide the courier with the latitude and longitude information to guide them to the accurate delivery or pickup address.

[0050] In some embodiments, to enrich the information recorded in the full address database and thus improve the success rate of retrieving latitude and longitude information based on the full address database, one or more embodiments of this specification also provide another full address database. That is, based on the full address database in the previous embodiment, in addition to the latitude and longitude information stored at the lowest level node, other levels also store matching latitude and longitude information. For example, the parent node of the country level also stores the latitude and longitude information matching the country, the node of the province level also stores the latitude and longitude information matching the province, and the node of the city level also stores the latitude and longitude information matching the city. This results in a full address database that records the mapping relationship between address entities at each level and their latitude and longitude information. In this way, during the application phase, even if the address text provided by the address provider has some missing entities, the matching latitude and longitude information can still be obtained based on the full address database, which can assist in location to a certain extent.

[0051] Because the full address database stores a large amount of data, the efficiency of retrieving latitude and longitude information based on the full address database is limited to some extent. Therefore, to further improve the retrieval efficiency of latitude and longitude information, in some embodiments, the address processing method provided in one or more embodiments of this specification may further include another latitude and longitude information retrieval scheme, namely, such as... Figure 3 As shown, Figure 3 This is a flowchart of another address processing method provided in an exemplary embodiment. After the step of obtaining the address text provided by the address provider, the address processing method provided in one or more embodiments of this specification may further include:

[0052] In step S210, a pre-established precise address database is invoked, and latitude and longitude information matching the address text is retrieved based on the precise address database; the confidence level of the mapping relationship between the address text and latitude and longitude information recorded in the precise address database is higher than a set confidence level threshold, and / or the number of times it is used is higher than a set number of times threshold, and the total number of mapping relationships recorded in the precise address database is less than the full address database.

[0053] Accordingly, step S220 or step S300 above is executed if no latitude and longitude information matching the address text is found in the accurate address database.

[0054] Understandably, during the process of building a full address database, mapping relationships with confidence levels higher than a set confidence threshold and / or usage counts higher than a set usage count threshold can be stored in another database to build a precise address database. It is evident that the precise address database typically records fewer mapping relationships than the full address database. The confidence and usage count thresholds can be set based on experience or actual needs, and this specification does not limit their specificity in the embodiments.

[0055] In the above, the confidence level of each mapping relationship can be obtained manually or determined based on the source of latitude and longitude information or the corresponding address level. For example, if the confidence level is determined based on the source of latitude and longitude information, the source can include the first latitude and longitude information selected by the user after locating the location on the map, the second latitude and longitude information detected by the courier's terminal device when receiving or picking up the package, and the third latitude and longitude information automatically matched from the map by the system based on the address text. Generally speaking, the first and second latitude and longitude information have higher confidence levels, so they can be considered to be higher than the set confidence threshold. Therefore, the mapping relationship between the first latitude and longitude information and its corresponding address text, as well as the mapping relationship between the second latitude and longitude information and its corresponding address text, can be recorded in the accurate address database. In addition, the confidence level of these mapping relationships can be marked as higher than the set confidence threshold in the full address database. For the third latitude and longitude information, the confidence level of its corresponding mapping relationship can be marked as lower than the set confidence threshold in the full address database. For example, when the confidence level is determined based on the address level corresponding to latitude and longitude information, it is understandable that since each address level is mapped to matching latitude and longitude information, the confidence level of the corresponding mapping relationship can be determined according to the level of the address level. The confidence level of the address level and the mapping relationship have an inverse trend. That is, the higher the address level (e.g., the national level is higher than the provincial level, the provincial level is higher than the municipal level, and the municipal level is higher than the district level), the lower the confidence level of the mapping relationship. The confidence level value of each level can be determined according to the preset mapping relationship between the level and the confidence level value.

[0056] Since the precise address database stores relatively few mapping relationships, in order to simplify the storage method and improve retrieval efficiency to a certain extent, the mapping relationship between address text and latitude and longitude information can be stored in a hierarchical manner instead of being divided into multiple levels of addresses for hierarchical mapping. In other words, instead of dividing the content contained in the address text into multiple levels of addresses for hierarchical mapping, the entire content contained in the address text can be directly mapped to latitude and longitude information.

[0057] Therefore, in order to improve retrieval efficiency and accuracy, step S210 is executed before step S300, that is, the precise address database is called first to retrieve latitude and longitude information that matches the address text based on the precise address database.

[0058] As can be seen, if latitude and longitude information matching the address text is found based on the precise address database, steps S210 and S300 can be skipped, reducing the computational burden of the retrieval. Conversely, if no latitude and longitude information matching the address text is found based on the precise address database, steps S210 and S300 are executed to perform a second retrieval based on the full address database, which improves the retrieval success rate.

[0059] As can be seen from the above, if no latitude and longitude information matching the address text is found in the accurate address database, executing step S210 will cause the second search to wait for the entity extraction model to output the address entity result, thus affecting the timeliness of the latitude and longitude information feedback. Therefore, in order to solve this technical problem, in some embodiments, steps S210 and S220 can be executed in parallel so that the address entity result can be obtained in a timely manner for the second search query when no latitude and longitude information matching the address text is found in the accurate address database.

[0060] For multilingual countries or regions, the same meaning in address text may be described in different languages. For example, in Southeast Asia, some numbers in address text may be described in Southeast Asian languages ​​or Arabic numerals. If a precise address database is queried based on this or directly input into an entity extraction model, it may result in the inability to retrieve matching latitude and longitude information based on the precise address database, or it may cause errors in the address entity results of the entity extraction model or prevent it from matching the address entities recorded in the full address database. Therefore, to solve this technical problem, in some embodiments, the address processing method provided by one or more embodiments of this specification further includes a scheme for preprocessing the address text. That is, after obtaining the address text provided by the address provider, the address processing method provided by one or more embodiments of this specification further includes:

[0061] In step S200, the non-standardized terms in the address text are corrected to standardized terms to obtain the preprocessed address text.

[0062] Accordingly, in steps S210 and S220, the address text used is the preprocessed address text.

[0063] Understandably, after obtaining the address text provided by the address provider, step S200 is executed first to preprocess the address text, thereby correcting non-standardized terms in the address text to standardized terms, including but not limited to: uniformly correcting the descriptions used to express numbers in the address text to Arabic numerals, and correcting the descriptions in the address text other than numbers that contain standard terms to standard terms.

[0064] Therefore, by preprocessing the address text, the address text can be standardized into terms that match the accurate address database and entity extraction model, which can better improve retrieval efficiency, retrieval accuracy, and model output accuracy.

[0065] In some embodiments, to further improve retrieval efficiency, the precise address database can be configured in memory; that is, the precise address database can be a memory address database.

[0066] Meanwhile, to balance performance and cost, in some embodiments, the full address database can be configured in non-memory environments, such as in a search service management platform like the Tisplus platform.

[0067] Since the mapping relationships stored in the full address database may not cover some areas, if the address text belongs to an uncovered area, matching latitude and longitude information cannot be retrieved based on the address text, which may affect subsequent logistics efficiency. Therefore, to solve this technical problem, in some embodiments, the address processing method provided in one or more embodiments of this specification may further include:

[0068] In step S420, if no latitude and longitude information matching the address entity result is found in the full address database, the map service is invoked and the address text is sent to the map service.

[0069] In step S510, upon receiving the latitude and longitude information returned by the map service based on the address text, the latitude and longitude information returned by the map service is fed back to the address provider.

[0070] Understandably, if no matching latitude and longitude information is found based on the full address database, it indicates that the full address database may not cover the area where the current address text is located. In this case, to ensure that latitude and longitude information with some reference value is returned, step S420 is executed to invoke a map service and send the address text to the map service. The map service may include, but is not limited to, at least one of the following: Gaode Map Service, Baidu Map Service, and Google Map Service.

[0071] After receiving the address text, the map service will query the matching latitude and longitude information based on the address text and provide feedback.

[0072] Therefore, upon receiving latitude and longitude information returned by the map service based on the address text, step S510 is executed to send the latitude and longitude information returned by the map service back to the address provider, which can also help the courier to successfully arrive at the delivery address or pickup address to a certain extent.

[0073] Upon receiving latitude and longitude information returned by a map service based on address text, in order to ensure that the full address database covers more areas and thus improve the success rate of subsequent latitude and longitude information retrieval, while avoiding the time consumption caused by calling the map service, and to better improve the feedback efficiency of latitude and longitude information, in some embodiments, the address processing method provided in one or more embodiments of this specification also proposes a full address database update scheme, that is: the address processing method may further include:

[0074] In step S520, the mapping relationship between the current address text and the matching latitude and longitude returned by the map service is recorded in the full address database to obtain the updated full address database.

[0075] In addition to the above-mentioned full address database update scheme, in some embodiments, the address processing method provided in one or more embodiments of this specification also proposes another full address database update scheme, that is, the address processing method may further include:

[0076] In step S610, according to the set update cycle, the pre-established logistics database is called, and the address table within the set time period before the current update cycle is read from the logistics database;

[0077] In step S620, the target latitude and longitude information of the same address text is determined based on the address text recorded in the address table and the initial latitude and longitude information of each address text.

[0078] In step S630, the full address database is updated according to the mapping relationship between the same address text and its target latitude and longitude information.

[0079] Understandably, steps S610 to S630 represent a scheme for maintaining the full address database in the background. During the maintenance of the full address database, steps S610 to S630 can be executed according to a set update cycle. The update cycle can be configured based on experience or actual needs; for example, it can be once a day, but it is not limited to this.

[0080] During step S610, the address table for a set time period prior to the current update cycle can be retrieved from the logistics database by accessing the logistics database. The address table may include a mapping relationship between the recipient address text and its initial latitude and longitude information, and a mapping relationship between the pickup address text and its initial latitude and longitude information. In some embodiments, to clearly distinguish between the recipient address and the pickup address and avoid confusion caused by the address table recording both addresses simultaneously, the address table may include a recipient address table and a pickup address table. The recipient address table records the mapping relationship between the recipient address text and its initial latitude and longitude information, and the pickup address table records the mapping relationship between the pickup address text and its initial latitude and longitude information.

[0081] In addition, since a full address database update is performed in each update cycle, in order to avoid wasting computing resources by obtaining duplicate data for updating in step S610, and to avoid missing storage mapping relationships due to incomplete data obtained in step S610, in some embodiments, the time period set before the current update cycle can be the time period between the current update cycle and the previous update cycle.

[0082] After reading the address table in step S610, step S620 is executed. First, the address texts in the address table are summarized, that is, address texts referring to the same address are merged into one address text, thus obtaining a single address text. Simultaneously, the initial latitude and longitude information corresponding to the address texts referring to the same address is associated with this single address text, thereby obtaining one or more initial latitude and longitude information for each single address text. For a single address text with one initial latitude and longitude information, this initial latitude and longitude information can be used as the target latitude and longitude information for that single address text. For a single address text with multiple initial latitude and longitude information, one or more embodiments of this specification provide corresponding target latitude and longitude information determination schemes for three scenarios, as follows:

[0083] The first scenario: Among the multiple initial latitude and longitude coordinates corresponding to an address, there is one initial latitude and longitude coordinate that appears most frequently.

[0084] In this case, step S620 above, determining the target latitude and longitude information of the same address text based on the address text recorded in the address table and the initial latitude and longitude information of each address text, may include:

[0085] In step S6211, based on the address text recorded in the address table and the initial latitude and longitude information of each address text, multiple initial latitude and longitude information corresponding to the same address text are obtained;

[0086] In step S6212, the initial latitude and longitude information that appears most frequently among the plurality of initial latitude and longitude information is taken as the target latitude and longitude information of the corresponding address text.

[0087] The technical principle of step S6211 can be found in the above description of step S620, and will not be repeated here.

[0088] After obtaining multiple initial latitude and longitude information corresponding to each text with the same address by executing step S6211, step S6212 is executed. According to the majority voting rule, the initial latitude and longitude information that appears most frequently among the multiple initial latitude and longitude information corresponding to each text with the same address is taken as the target latitude and longitude information of the corresponding address text.

[0089] For example, suppose there are five initial latitude and longitude coordinates corresponding to the same address text, namely Lat1, Lat2, Lat1, Lat1, and Lat3. It can be seen that Lat1 appears the most frequently, so the target latitude and longitude coordinates corresponding to the same address text are Lat1.

[0090] If the initial latitude and longitude information with the highest frequency does not exist during the execution of step S6212, the target latitude and longitude information determination scheme under the second or third scenario can be adopted as appropriate.

[0091] The second scenario: Among the multiple initial latitude and longitude information corresponding to the same address text, there is no initial latitude and longitude information that appears most frequently, and the multiple initial latitude and longitude information corresponding to the same address text fall into the same lowest level address region.

[0092] In this case, step S620 above, which involves determining the target latitude and longitude information of the same address text based on the address text recorded in the address table and the initial latitude and longitude information of each address text, may further include:

[0093] In step S6221, when the lowest level address regions corresponding to the plurality of initial latitude and longitude information are all the same, the first centroid of the plurality of initial latitude and longitude information is calculated.

[0094] In step S6222, the initial latitude and longitude information closest to the first centroid among the plurality of initial latitude and longitude information is used as the target latitude and longitude information of the corresponding address text.

[0095] Understandably, during the execution of step S6212, if it is found that there is no initial latitude and longitude information with the highest frequency, it can be determined whether multiple initial latitude and longitude information fall into the same lowest-level address region. For example, since each initial latitude and longitude information can be used to locate the corresponding lowest-level address region, such as the district level or road level, taking the district level as an example, if multiple initial latitude and longitude information fall into the same district, then step S6221 is executed to calculate the first centroid of multiple initial latitude and longitude information. The calculation principle can be found in relevant technologies and will not be explained in detail here.

[0096] After obtaining the first centroid, step S6222 is executed to select the one that is closest to the first centroid from multiple initial latitude and longitude information, and use it as the target latitude and longitude information of the corresponding address text.

[0097] Conversely, if multiple initial latitude and longitude coordinates fall into different zones, the scheme described in the third scenario will be executed.

[0098] The third scenario: Among the multiple initial latitude and longitude information corresponding to the same address text, there is no initial latitude and longitude information that appears most frequently, and the multiple initial latitude and longitude information corresponding to the same address text fall into different lowest-level address regions.

[0099] In this case, step S620 above, which involves determining the target latitude and longitude information of the same address text based on the address text recorded in the address table and the initial latitude and longitude information of each address text, may further include:

[0100] In step S6231, when the lowest level address regions corresponding to the multiple initial latitude and longitude information are partially different, the second centroid of the initial latitude and longitude information contained in the lowest level address region that appears most frequently is calculated.

[0101] In step S6232, the initial latitude and longitude information closest to the second centroid among the initial latitude and longitude information contained in the lowest level address region that appears most frequently is taken as the target latitude and longitude information of the corresponding address text.

[0102] Understandably, during step S6212, if the highest-frequency initial latitude and longitude information is not found, it can be determined whether multiple initial latitude and longitude information fall into the same lowest-level address region. For example, since each initial latitude and longitude information can locate the corresponding lowest-level address region, such as the district level or road level, taking the district level as an example, if multiple initial latitude and longitude information do not fall into the same district, then step S6231 is executed. First, it counts which districts the multiple initial latitude and longitude information fall into, and then identifies which district contains the most initial latitude and longitude information. This gives the lowest-level address region with the highest frequency. For example, assuming there are 5 initial latitude and longitude information, 3 fall into district L1 and the other 2 fall into district L2, then according to the majority voting rule, district L1 is the lowest-level address region with the highest frequency. Next, the second centroid of the initial latitude and longitude information contained in district L1 is calculated.

[0103] After obtaining the second centroid, step S6232 is executed. Continuing with the above example, the latitude and longitude information contained in region L1 that is closest to the second centroid is used as the target latitude and longitude information of the corresponding address text.

[0104] After obtaining each identical address text and its corresponding target latitude and longitude information through any of the above embodiments, step S630 is executed to update the full address database using the obtained mapping relationship. During the update of the full address database, existing mapping relationships do not need to be stored again; however, the usage count of the mapping relationship can be updated. For example, assuming that the mapping relationship is recorded n times in the address table and m times in the full address database during this update, the usage count of the mapping relationship can be updated to m+n.

[0105] Therefore, by dynamically updating the full address database, the database can be gradually improved, covering logistics address information and latitude and longitude information in more regions. This will lay a solid foundation for logistics information in countries with poor basic logistics services, and thus better solve the problem of latitude and longitude matching.

[0106] Furthermore, to improve the mapping coverage of the precise address database and thus enhance the accuracy and retrieval efficiency of latitude and longitude information, one or more embodiments of this specification provide two schemes for updating the precise address database:

[0107] The first precise address database update scheme: update based on the full address database.

[0108] Based on this, the address processing method provided in one or more embodiments of this specification may further include:

[0109] In step S711, according to the set update cycle, the full address database is called, and target mapping relationships with a confidence level higher than the set confidence level threshold and / or a number of uses higher than the set number of uses threshold are selected from the full address database;

[0110] In step S712, the precise address database is updated according to the target mapping relationship.

[0111] As described above, the full address database records the usage frequency and confidence level of each mapping relationship. Therefore, steps S711 and S712 can be executed according to the set update cycle to update the accurate address database in a timely manner. The set update cycle can be the same as that of the full address database, but to ensure the integrity of the obtained target mapping relationships, steps S711 and S712 can be executed after each update of the full address database.

[0112] During the execution of step S711, the full address database is first invoked to select target mapping relationships that are not recorded in the precise address database and whose confidence level and / or number of uses meet the set requirements.

[0113] After obtaining the target mapping relationship, step S712 is executed to record the target mapping relationship into the precise address database, so as to update the precise address database.

[0114] The second precise address database update scheme is to update based on error correction feedback information.

[0115] Based on this, the address processing method provided in one or more embodiments of this specification may further include:

[0116] In step S721, upon receiving a feedback message indicating an incorrect delivery address, a pre-established logistics database is invoked, and the target address text corresponding to the incorrect delivery address and its correct latitude and longitude information are selected from the logistics database.

[0117] In step S722, the accurate address database is updated based on the target address text and its correct latitude and longitude information.

[0118] Understandably, upon receiving a feedback message indicating an incorrect delivery address, step S721 is executed to invoke a pre-established logistics database and select the target address text corresponding to the incorrect delivery address and its correct latitude and longitude information from the logistics database.

[0119] The aforementioned feedback message can be received by the logistics system. Understandably, the logistics system's database records the receipt and pickup records for each express delivery order. For example, if an express delivery order was originally assigned to a courier responsible for area E1 but was ultimately reassigned to a courier responsible for area E2, the receipt record includes the courier reassignment information. Therefore, the logistics system can determine whether there are records related to courier reassignment information based on the receipt and pickup records in the logistics database. If so, the logistics system will generate the aforementioned feedback message and send it to the execution entity that performs the address processing method provided in one or more embodiments of this specification.

[0120] Upon receiving the above feedback message, step S721 is executed to access the logistics database of the logistics system, and to select the relevant records containing courier re-delivery information from the receipt and pickup records of the logistics database. Then, the target address text and its correct latitude and longitude information are extracted from the relevant records.

[0121] Next, step S722 is executed to update the corresponding mapping relationship in the precise address database based on the target address text and its correct latitude and longitude information, thereby realizing the update of the precise address database.

[0122] Therefore, by dynamically updating the accurate address database, it is possible to gradually improve the database, enabling countries with poor basic logistics services to also have a certain amount of accurate mapping relationships, thus laying a better foundation for logistics services.

[0123] Figure 4 This is a schematic structural diagram of a device provided in an exemplary embodiment. For example... Figure 4 As shown, device 400 mainly consists of a communication interface 402, a user interface 404, a processor 406, and a data storage 408. These components are interconnected and communicate with each other via a system bus, network, or other connection mechanism 410. The communication interface 402 enables device 400 to communicate with other devices, access networks, and transmission networks via analog or digital modulation. For example, the communication interface 402 may include a chipset and antenna for wireless communication with a radio access network or access point. Furthermore, the communication interface 402 can be a wired interface such as Ethernet, Token Ring, or a USB port, or a wireless interface such as Wi-Fi, Bluetooth, Global Positioning System (GPS), or a wide-area wireless interface (e.g., WiMAX or LTE). Of course, the communication interface 402 can also support other forms of physical layer interfaces and standard or proprietary communication protocols. The communication interface 402 may also include multiple physical communication interfaces, such as Wi-Fi, Bluetooth, and wide-area wireless interfaces.

[0124] User interface 404 includes receiving user input and providing output to the user. Therefore, user interface 404 may include input components such as a keypad, keyboard, touch-sensitive or presence-sensitive panel, computer mouse, trackball, joystick, microphone, still camera, and video camera, and output components such as a display screen (which may be combined with a touch-sensitive panel), CRT, LCD, LED, display using DLP technology, printer, and other similar devices known or developed in the future. User interface 404 may also generate auditory output via speakers, speaker jacks, audio output ports, audio output devices, headphones, and other similar devices known or developed in the future. In some embodiments, user interface 404 may include software, circuitry, or other forms of logic capable of transmitting and receiving data from external user input / output devices. Additionally or alternatively, device 400 may support remote access from other devices via communication interface 402 or another physical interface (not shown). User interface 404 may be configured to receive user input, the position and movement of which may be indicated by indicators or cursors described herein. User interface 404 may also be configured as a display device for rendering or displaying text fragments.

[0125] Processor 406 may contain one or more general-purpose processors and / or special-purpose processors.

[0126] Data storage 408 may include one or more volatile and / or non-volatile storage components and may be integrated wholly or partially with processor 406. Data storage 408 may include removable and non-removable components.

[0127] Processor 406 is capable of executing program instructions 418 (e.g., compiled or uncompiled program logic and / or machine code) stored in data storage 408 to perform the various functions described herein. Data storage 408 may contain a non-transitory computer-readable medium on which program instructions are stored, which, when executed by device 400, enable device 400 to perform any methods, processes, or functions disclosed in this specification and / or the accompanying drawings. Processor 406 executing program instructions 418 may result in processor 406 using data 412.

[0128] For example, program instructions 418 may include an operating system 422 (e.g., an operating system kernel, device drivers, and / or other modules) installed on device 400 and one or more applications 420 (e.g., a browser, social application, or game application). Similarly, data 412 may include operating system data 416 and application data 414. Operating system data 416 is primarily accessible to the operating system 422, while application data 414 is primarily accessible to one or more applications 420. Application data 414 may reside in a file system visible or hidden from the user of device 400.

[0129] Application 420 can communicate with operating system 422 through one or more application programming interfaces (APIs). These APIs help application 420 read and / or write application data 414, transmit or receive information via communication interface 402, receive or display information on user interface 404, etc.

[0130] In some terminology, application 420 may be simply referred to as "app". Furthermore, application 420 can be downloaded to device 400 through one or more online app stores or app markets. However, applications can also be installed on device 400 in other ways, such as through a web browser or a physical interface on device 400 (e.g., a USB port).

[0131] Please refer to Figure 5 , Figure 5 This is a block diagram of an address processing apparatus 500 provided in an exemplary embodiment, which can be applied to, for example... Figure 4 The device shown implements the technical solution of this specification. The address processing device 500 may include:

[0132] The acquisition module 510 is configured to: in response to the address processing request of the address provider, acquire the address text provided by the address provider;

[0133] The extraction module 520 is configured to: call a pre-established model library, select an entity extraction model from the model library that matches the regional information recorded in the address text, and input the address text into the entity extraction model to obtain address entity results; the address entity results include address entities corresponding to different levels of addresses, and the address granularity of different levels of addresses is different;

[0134] The retrieval module 530 is configured to: call a pre-established full address database and retrieve latitude and longitude information that matches the address entity result based on the full address database;

[0135] Feedback module 540 is configured to: provide the retrieved latitude and longitude information to the address provider.

[0136] In some embodiments, the retrieval module 530 can also be configured as:

[0137] Before calling the pre-established full address database, a pre-established precise address database is called, and latitude and longitude information matching the address text is retrieved based on the precise address database; the confidence level of the mapping relationship between the address text and latitude and longitude information recorded in the precise address database is higher than a set confidence level threshold, and / or the number of times it is used is higher than a set number of times threshold, and the total number of the mapping relationships recorded in the precise address database is less than that in the full address database;

[0138] Correspondingly, the retrieval module 530 calls the pre-established full address database and performs the operation of retrieving latitude and longitude information that matches the address entity result based on the full address database. This operation is performed if no latitude and longitude information matching the address text is found based on the precise address database.

[0139] In some embodiments, the address processing device 500 may further include:

[0140] The full address database update module is configured as follows:

[0141] According to the set update cycle, the pre-established logistics database is called, and the address table within the set time period before the current update cycle is read from the logistics database;

[0142] Based on the address texts recorded in the address table and the initial latitude and longitude information of each address text, determine the target latitude and longitude information of the same address text;

[0143] The full address database is updated based on the mapping relationship between the same address text and its target latitude and longitude information.

[0144] In some embodiments, the process by which the full address database update module determines the target latitude and longitude information of the same address text based on the address text recorded in the address table and the initial latitude and longitude information of each address text is configured as follows:

[0145] Based on the address text recorded in the address table and the initial latitude and longitude information of each address text, multiple initial latitude and longitude information corresponding to the same address text are obtained;

[0146] The initial latitude and longitude information that appears most frequently among the multiple initial latitude and longitude information is used as the target latitude and longitude information of the corresponding address text.

[0147] In some embodiments, the process by which the full address database update module determines the target latitude and longitude information of the same address text based on the address text recorded in the address table and the initial latitude and longitude information of each address text is further configured as follows:

[0148] If there is no initial latitude and longitude information with the highest frequency among the multiple initial latitude and longitude information, and the lowest level address region corresponding to each of the multiple initial latitude and longitude information is the same, calculate the first centroid of the multiple initial latitude and longitude information.

[0149] The initial latitude and longitude information that is closest to the first centroid among the multiple initial latitude and longitude information is used as the target latitude and longitude information of the corresponding address text.

[0150] In some embodiments, the process by which the full address database update module determines the target latitude and longitude information of the same address text based on the address text recorded in the address table and the initial latitude and longitude information of each address text is further configured as follows:

[0151] If there is no initial latitude and longitude information with the highest frequency among the multiple initial latitude and longitude information, and the lowest level address regions corresponding to each of the multiple initial latitude and longitude information are partially different, calculate the second centroid of the initial latitude and longitude information contained in the lowest level address region with the highest frequency.

[0152] The initial latitude and longitude information that is closest to the second centroid among the initial latitude and longitude information contained in the lowest level address region that appears most frequently is taken as the target latitude and longitude information of the corresponding address text.

[0153] In some embodiments, the address processing device 500 may further include:

[0154] The precise address database update module is configured as follows:

[0155] According to the set update cycle, the full address database is called, and target mapping relationships with a confidence level higher than the set confidence threshold and / or a usage frequency higher than the set frequency threshold are selected from the full address database;

[0156] The precise address database is updated based on the target mapping relationship.

[0157] In some embodiments, the precise address database update module can also be configured as follows:

[0158] Upon receiving a feedback message indicating an incorrect delivery address, the system invokes a pre-established logistics database and selects the target address text corresponding to the delivery address error and its correct latitude and longitude information from the database.

[0159] The accurate address database is updated based on the target address text and its correct latitude and longitude information.

[0160] In some embodiments, the retrieval module 530 can also be configured as:

[0161] If no latitude and longitude information matching the address entity result is found in the full address database, the map service is invoked and the address text is sent to the map service.

[0162] Upon receiving latitude and longitude information returned by the map service based on the address text, the latitude and longitude information returned by the map service is fed back to the address provider.

[0163] In some embodiments, the entity extraction model is trained based on a pre-established address training set, which includes multiple sets of training samples. Each set of training samples includes training address text and entity labels that match the training address text. The entity labels include address entities corresponding to the different levels of addresses.

[0164] In some embodiments, an entity extraction model is trained on a training set of addresses whose address format similarity meets a set similarity threshold.

[0165] For ease of description, the above devices are described by dividing them into various modules or units based on their functions. Of course, when implementing one or more of these specifications, the functions of each module or unit can be implemented in the same or different software and / or hardware, or a module that performs the same function can be implemented by a combination of multiple sub-modules or sub-units, etc. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed.

[0166] Based on the same concept as the methods described above, this specification also provides an electronic device, including: a processor; and a memory for storing processor-executable instructions; wherein the processor performs the steps of the method as described in any of the above embodiments by executing the executable instructions.

[0167] Based on the same concept as the methods described above, this specification also provides a computer-readable storage medium having computer instructions stored thereon that, when executed by a processor, implement the steps of the methods as described in any of the above embodiments.

[0168] Based on the same concept as the methods described above, this specification also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the methods as described in any of the above embodiments.

[0169] What those skilled in the art will understand is:

[0170] In this specification, the terms "comprising," "including," or any other variations thereof are intended to cover a non-exclusive inclusion, such that a process, method, product, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, product, or apparatus. Without further limitation, the presence of additional identical or equivalent elements in a process, method, product, or apparatus that includes said elements is not excluded.

[0171] In this specification, “a,” “an,” and “the” do not specifically refer to the singular, but may also include the plural.

[0172] In this specification, ordinal numbers such as "first," "second," etc., do not necessarily indicate order; they are often used to distinguish between objects. For example, "first server" and "second server" usually refer to two servers. To differentiate between these two servers, they are described as "first server" and "second server." Of course, sometimes these two servers may be the same server.

[0173] In this specification, unless explicitly stated otherwise, "receiving and sending data" does not necessarily mean direct receiving and sending; it can also mean indirect receiving and sending. For example, A receiving data sent by B can be understood as A directly receiving the data sent by B, or it can be understood as A indirectly receiving the data sent by B through other entities such as C. Similarly, B sending data to A can be understood as B sending the data directly to A, or it can be understood as B indirectly sending the data to A through other entities such as C. Here, C can be one entity, or it can be two or more entities.

[0174] In this specification, unless explicitly stated otherwise, the relationships between structures can be direct or indirect. For example, when describing "A is connected to B," unless it is explicitly stated that A and B are directly connected, it should be understood that A can be directly connected to B or indirectly connected to B. Similarly, when describing "A is on top of B," unless it is explicitly stated that A is directly above B (AB is adjacent and A is above B), it should be understood that A can be directly above B or indirectly above B (AB is separated by other elements, and A is above B). And so on.

[0175] This specification uses specific terms to describe embodiments thereof. Terms such as "an embodiment," "one embodiment," and / or "some embodiments" refer to a particular feature, structure, or characteristic associated with at least one embodiment of this specification. Therefore, it should be emphasized and noted that references to "an embodiment," "one embodiment," or "an alternative embodiment" in different locations throughout this specification do not necessarily refer to the same embodiment. Furthermore, those skilled in the art can combine and integrate the different embodiments or examples described herein, as well as the features of those different embodiments or examples, without contradiction.

[0176] Although one or more embodiments of this specification provide method steps as described in the embodiments or flowcharts, it is understood that the order of steps listed in the embodiments or flowcharts is only one of many possible execution orders and does not represent the only execution order. Therefore, when the claims involve method steps, any changes or adjustments to the order of such steps, or the parallelism between steps, are also within the scope of protection of the claims.

Claims

1. An address processing method, comprising: In response to the address processing request from the address provider, obtain the address text provided by the address provider; A pre-established model library is invoked, and an entity extraction model that matches the regional information recorded in the address text is selected from the model library. The address text is then input into the entity extraction model to obtain the address entity results. The address entity results include address entities corresponding to different levels of addresses, and the address granularity of different levels of addresses is different. Call the pre-established full address database and retrieve latitude and longitude information that matches the address entity result based on the full address database; The retrieved latitude and longitude information is fed back to the address provider.

2. The method according to claim 1, after the step of obtaining the address text provided by the address provider, the method further includes: Call a pre-established precise address database and retrieve latitude and longitude information that matches the address text based on the precise address database; The confidence level of the mapping relationship between the address text and latitude and longitude information recorded in the precise address database is higher than a set confidence level threshold, and / or the number of times it is used is higher than a set number of times threshold, and the total number of the mapping relationships recorded in the precise address database is less than the total number of addresses in the full address database. The step of calling the pre-established model library, or the step of calling the pre-established full address library, is performed when no latitude and longitude information matching the address text is found in the precise address library.

3. The method according to claim 1, further comprising: According to the set update cycle, the pre-established logistics database is called, and the address table within the set time period before the current update cycle is read from the logistics database; Based on the address texts recorded in the address table and the initial latitude and longitude information of each address text, determine the target latitude and longitude information of the same address text; The full address database is updated based on the mapping relationship between the same address text and its target latitude and longitude information.

4. The method according to claim 3, wherein the step of determining the target latitude and longitude information of the same address text based on the address text recorded in the address table and the initial latitude and longitude information of each address text includes: Based on the address text recorded in the address table and the initial latitude and longitude information of each address text, multiple initial latitude and longitude information corresponding to the same address text are obtained; The initial latitude and longitude information that appears most frequently among the multiple initial latitude and longitude information is used as the target latitude and longitude information of the corresponding address text.

5. The method according to claim 4, wherein if no initial latitude and longitude information appears most frequently among the plurality of initial latitude and longitude information, the step of determining the target latitude and longitude information of the same address text based on the address text recorded in the address table and the initial latitude and longitude information of each address text further includes: When the lowest-level address regions corresponding to the multiple initial latitude and longitude information are all the same, calculate the first centroid of the multiple initial latitude and longitude information. The initial latitude and longitude information that is closest to the first centroid among the multiple initial latitude and longitude information is used as the target latitude and longitude information of the corresponding address text.

6. The method according to claim 4 or 5, wherein if no initial latitude and longitude information appears most frequently among the plurality of initial latitude and longitude information, the step of determining the target latitude and longitude information of the same address text based on the address text recorded in the address table and the initial latitude and longitude information of each address text further includes: When the lowest-level address regions corresponding to the multiple initial latitude and longitude information are partially different, the second centroid of the initial latitude and longitude information contained in the lowest-level address region that appears most frequently is calculated. The initial latitude and longitude information that is closest to the second centroid among the initial latitude and longitude information contained in the lowest level address region that appears most frequently is taken as the target latitude and longitude information of the corresponding address text.

7. The method according to claim 2, further comprising: According to the set update cycle, the full address database is called, and target mapping relationships with a confidence level higher than the set confidence threshold and / or a usage frequency higher than the set frequency threshold are selected from the full address database; The precise address database is updated based on the target mapping relationship.

8. The method according to claim 2 or 7, further comprising: Upon receiving a feedback message indicating an incorrect delivery address, the system invokes a pre-established logistics database and selects the target address text corresponding to the delivery address error and its correct latitude and longitude information from the database. The accurate address database is updated based on the target address text and its correct latitude and longitude information.

9. The method according to claim 1 or 2, further comprising: If no latitude and longitude information matching the address entity result is found in the full address database, the map service is invoked and the address text is sent to the map service. Upon receiving latitude and longitude information returned by the map service based on the address text, the latitude and longitude information returned by the map service is fed back to the address provider.

10. The method according to claim 1, wherein the entity extraction model is trained based on a pre-established address training set, the address training set including multiple sets of training samples, each set of training samples including training address text and entity labels matching the training address text; the entity labels include address entities corresponding to the different level addresses.

11. The method according to claim 10, wherein an entity extraction model is trained based on an address training set whose address format similarity satisfies a set similarity threshold.

12. An electronic device, comprising: processor; A memory for storing processor-executable instructions; wherein the processor implements the steps of the method as described in any one of claims 1 to 11 by executing the executable instructions.

13. A computer-readable storage medium having stored thereon computer instructions that, when executed by a processor, implement the steps of the method as claimed in any one of claims 1 to 11.

14. A computer program product comprising a computer program that, when executed by a processor, implements the steps of the method as described in any one of claims 1 to 11.

Images