Address supplement method and device, equipment and storage medium

Abstract

The embodiment of the invention provides an address supplementing method and device, equipment and a storage medium. In the embodiment of the invention, a search module can be controlled to perform network search on address keywords in text information, and a search result containing the address keywords is provided for a first model; based on first task description information contained in a preset cue word, driving a first model to supplement the address keyword based on the search result so as to generate address information corresponding to the address keyword; and constructing an address supplement result for the text information according to the generated address information. The information in the networking environment has comprehensiveness and real-time performance, and the address keyword is searched in the networking environment through the search module, so that the latest and comprehensive search result containing the address keyword can be searched. On this basis, the first model can more accurately and comprehensively supplement the address keyword based on the search result, so that more accurate and complete address information can be generated, and the accuracy of address supplement is improved.

Description

Technical Field

[0001] The present application relates to the field of information processing technology, and in particular to an address supplement method, device, equipment and storage medium. Background Art

[0002] With the popularity of e-commerce, food delivery, home housekeeping and other services, user addresses have become the basis for accurate and efficient delivery. Whether it is online shopping, food delivery or home service, the platform needs to know the user's address to ensure that the goods or services can be delivered accurately and in a timely manner. However, the user address provided by the user is often incomplete, making it impossible for the platform to obtain a complete and accurate address, and thus unable to deliver goods or services on time.

[0003] At present, the address supplement scheme adopted is usually to search for a sub-address matching the user address provided by the user from the knowledge base, and based on the found sub-address, supplement the user address in the text information to generate the user's detailed address. However, in this address supplement scheme, the sub-addresses included in the knowledge base are not comprehensive, and it may be impossible to supplement the user address with a complete detailed address, and the accuracy of address supplement is poor. Summary of the invention

[0004] Multiple aspects of the present application provide an address supplement method, apparatus, device and storage medium to improve the accuracy of address supplement.

[0005] The present application embodiment provides an address supplement method, which is applicable to an address supplement device, and the method includes:

[0006] Get the text information entered by the user;

[0007] performing an address extraction operation on the text information to extract address keywords from the text information;

[0008] Controlling a preset search module to perform a network search according to the address keyword to obtain search results containing the address keyword;

[0009] Controlling the search module to provide the search results to the first model;

[0010] Based on the first task description information included in the preset prompt word, driving the first model to supplement the address keyword based on the search result to generate address information corresponding to the address keyword;

[0011] An address supplement result is constructed for the text information according to the address information generated by the first model.

[0012] The embodiment of the present application also provides an address supplement device, the address supplement device comprising an extraction module, a search module and a supplement module;

[0013] The extraction module is used to obtain text information input by the user and perform an address extraction operation on the text information to extract address keywords from the text information;

[0014] The search module is used to perform a network search based on the address keyword to obtain search results containing the address keyword, and provide the search results to the supplement module;

[0015] The supplementing module is used to drive the first model to supplement the address keyword based on the search results based on the first task description information contained in the preset prompt word to generate address information corresponding to the address keyword; and construct an address supplement result for the text information based on the address information generated by the first model.

[0016] The embodiment of the present application also provides a computing device, including: a memory, a processor, and a communication component;

[0017] The memory is used to store one or more computer instructions;

[0018] The processor is coupled to the memory and the communication component, and is configured to execute the one or more computer instructions to perform the aforementioned address supplementation method.

[0019] An embodiment of the present application also provides a computer-readable storage medium storing computer instructions. When the computer instructions are executed by one or more processors, the one or more processors are caused to execute the aforementioned address supplement method.

[0020] An embodiment of the present application also provides a computer program product, including a computer program; when the computer program is executed by a processor, the processor is caused to execute the aforementioned address supplement method.

[0021] In an embodiment of the present application, the address supplement device includes a search module that supports network search, which can extract address keywords from the text information input by the user, control the preset search module, and perform network search based on the address keywords to obtain search results containing the address keywords; control the search module to provide the search results to the first model, and based on the first task description information contained in the preset prompt word, drive the first model to supplement the address keywords based on the search results to generate address information corresponding to the address keywords, and construct address supplement results for the text information based on the address information generated by the first model. Accordingly, the address keywords are searched in a networked environment through the search module, and the networked environment can ensure the comprehensiveness and real-time nature of the information related to the address keywords, so that the latest and comprehensive search results containing the address keywords can be searched. On this basis, the first model can supplement the address keywords more accurately and comprehensively based on the search results, thereby making the generated address information corresponding to the address keywords more accurate and complete, thereby improving the accuracy of address supplementation. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0023] Figure 1 A flowchart of an address supplement method provided by an exemplary embodiment of the present application;

[0024] Figure 2 A schematic diagram of the structure of a preset knowledge base provided for an exemplary embodiment of the present application;

[0025] Figure 3 A flowchart of an application scenario provided by an exemplary embodiment of the present application;

[0026] Figure 4 A schematic structural diagram of an address supplement device provided by another exemplary embodiment of the present application;

[0027] Figure 5 A schematic diagram of the structure of a computing device is provided for yet another exemplary embodiment of the present application. DETAILED DESCRIPTION

[0028] In order to make the purpose, technical solution and advantages of the present application clearer, the technical solution of the present application will be clearly and completely described below in combination with the specific embodiments of the present application and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present application.

[0029] At present, when providing corresponding services to ordering users, service platforms (such as e-commerce platforms and housekeeping platforms) need to obtain the detailed address of the ordering user. However, when the ordering user provides the address to the service platform, the address provided is often incomplete, making it impossible for the service platform to obtain the accurate address of the ordering user. In response to this situation, the service platform usually searches for a sub-address that matches the user address contained in the text information input by the ordering user from a preset knowledge base, and supplements the user address in the text information based on the sub-address found. However, the sub-addresses contained in the knowledge base are not comprehensive, and it may be impossible to supplement the user address with a complete detailed address, and the accuracy of the address supplement is poor.

[0030] To improve this problem, the embodiment of the present application provides a solution, the basic idea is: to design an address supplement device including a search module, the search module can support network search for address keywords. The address supplement device can extract address keywords from the text information input by the user, and search for address keywords in a networked environment through the search module. The networked environment can ensure the comprehensiveness and real-time nature of information related to address keywords, so that the latest and comprehensive search results containing address keywords can be searched. On this basis, based on the first task description information contained in the preset prompt word, the first model can be driven to perform a more accurate and comprehensive supplement to the address keywords based on the search results, thereby making the generated address information corresponding to the address keywords more accurate and complete, thereby improving the accuracy of address supplementation.

[0031] The technical solutions provided by various embodiments of the present application are described in detail below in conjunction with the accompanying drawings.

[0032] Figure 1 A flowchart of an address supplement method provided by an exemplary embodiment of the present application. The method can be executed by an address supplement device, which can be implemented as a combination of software and / or hardware, and the data processing device can be integrated in a computing device. From the perspective of physical devices, the method can be applied to the server. Figure 1 , the method comprising:

[0033] Step 100, obtaining text information input by the user;

[0034] Step 101, performing an address extraction operation on the text information to extract address keywords from the text information;

[0035] Step 102, controlling a preset search module to perform a network search according to the address keyword to obtain search results containing the address keyword;

[0036] Step 103, controlling the search module to provide the search results to the first model;

[0037] Step 104, based on the first task description information included in the preset prompt word, driving the first model to supplement the address keyword based on the search results to generate address information corresponding to the address keyword;

[0038] Step 105: construct an address supplement result for the text information according to the address information generated by the first model.

[0039] The address supplement method provided in this embodiment can be applied to various scenarios where the address needs to be automatically supplemented, for example, automatically supplementing the delivery address entered by the user on the e-commerce platform, automatically supplementing the door-to-door service address entered by the user on the housekeeping platform, extracting the address in the chat service, etc. This embodiment does not limit the application scenario.

[0040] In step 100, text information input by the user may be obtained. The text information is the information to be supplemented for the address, and the text information may include but is not limited to name, address, contact number, etc. The address supplement device may be deployed on a service platform, and the service platform may support multiple business functions. In response to a text information submission operation initiated by the user for any business function, the text information input by the user may be sent to the address supplement device, and the content of the text information is associated with the business function.

[0041] After obtaining the text information, an address extraction operation may be performed on the text information to extract address keywords from the text information. The number of address keywords extracted from the text information may be one or more. Address keywords refer to address-related information in the text information. These address-related information are usually composed of a series of descriptive elements, which together constitute an accurate location identifier. These elements may include but are not limited to country, region, province, city, town, street, house number, etc. If the address keyword cannot be extracted from the text information, a prompt message may be generated to indicate that the text information does not contain the address keyword and the text information is ignored.

[0042] There are many ways to implement the extraction of address keywords, and the specific implementation details will be described in detail later and will not be elaborated here.

[0043] Continue to refer Figure 1In step 102, a preset search module may be controlled to perform a network search based on the address keyword to obtain search results containing the address keyword. The search module allows the user to directly access a specific search engine or search tool through a browser to quickly obtain the required information. The search module may be an online search plug-in, a map search plug-in, etc. These plug-ins integrate the powerful functions of the search engine and provide additional convenience and customization.

[0044] In step 103 to step 104, the search module can be controlled to provide the search results to the first model, and based on the first task description information contained in the preset prompt word, the first model is driven to supplement the address keyword based on the search results to generate address information corresponding to the address keyword. The first model is a large model that can supplement the address of the acquired address keyword based on the search results. The large model refers to a machine learning model with a large number of parameters and complex structures. These models can be applied to process large-scale data and complex problems.

[0045] It can be understood that the address supplement device includes a search module and a first model, and the address supplement device can provide the search results obtained by the search module to the first model, so as to use the search results as input of the first model. In this embodiment, the first model is also configured with a preset prompt word, and the prompt word can include multiple task description information, and the multiple task description information is used to drive the large model to perform multiple tasks. The first task description information included in the prompt word can be used to guide the first model to supplement the address of the acquired address keywords according to the search results provided by the search module, so as to generate address information corresponding to the address keywords. The first task description information can also limit the output format of the supplemented address information, for example, JSON format, XML format, etc.

[0046] In step 105, an address supplement result may be constructed for the text information according to the address information generated by the first model. The address information generated here is the output obtained by the first model using the search results as input. Based on this, the address supplement device may construct an address supplement result for the text information based on the output of the first model. The address supplement result may include address information, other information in the text information except address keywords, etc. After the address supplement result is constructed for the text information, the address supplement result may be directly provided to the user who inputs the text information, or the address supplement result may be provided to a service platform deployed with the address supplement device, so that the service platform may use the address supplement result to perform business functions.

[0047] In summary, in this embodiment, address keywords can be extracted from the text information input by the user, and the preset search module can be controlled to perform a network search based on the address keywords to obtain search results containing the address keywords; the search module is controlled to provide the search results to the first model, and based on the first task description information contained in the preset prompt word, the first model is driven to supplement the address keywords based on the search results to generate address information corresponding to the address keywords, and the address supplement results are constructed for the text information based on the address information generated by the first model. Accordingly, the address keywords are searched in a networked environment through the search module, and the networked environment can ensure the comprehensiveness and real-time nature of the information related to the address keywords, so that the latest and comprehensive search results containing the address keywords can be searched. On this basis, the first model can supplement the address keywords more accurately and comprehensively based on the search results, thereby making the generated address information corresponding to the address keywords more accurate and complete, thereby improving the accuracy of address supplementation.

[0048] In the above or below embodiments, various implementations may be used to control a preset search module to perform a network search based on address keywords to obtain search results.

[0049] In an optional implementation, the search module can be controlled to construct a query request based on the address keyword, and the query request is sent to a search engine associated with the search module through the search module. The query request includes the address keyword. As mentioned above, the search module can be an online search plug-in, a map search plug-in, etc. These search plug-ins are pre-configured with one or more search engines, which can be general search engines or search engines in specific fields (such as map search engines, etc.).

[0050] On this basis, the search module can be used to filter the page information returned by the search request in response to the query request to delete the page information that does not contain the address keyword, and the remaining page information after filtering is used as the search result. For example, when searching in the search engine, some page information such as advertisements and promotional articles may be searched out. These page information do not contain the address keyword and are page information unrelated to the address keyword. Therefore, these page information that does not contain the address keyword can be deleted to filter the page information returned by the search engine. Of course, it is also possible to delete the page information that contains the address keyword but does not involve other address content from the page information returned by the search engine. For example, if the page information contains the address keyword "Chaoyang", but the address keyword is only "Chaoyang resident Li" in the page information, and other address content related to "Chaoyang" is not involved, the page information can be deleted.

[0051] Of course, the search engine can also perform filtering of page information. The search engine's URL and search parameters (such as sorting method, filtering conditions, etc.) can be carried in the query request, so that the search engine can filter the searched page information according to the filtering conditions and directly return the filtered page information as search results to the search module. In this case, the search module does not need to perform repeated filtering.

[0052] In actual applications, by searching the address keywords online, the latest and comprehensive search results containing the address keywords can be found. In addition, by filtering the search results, information irrelevant to address supplementation can be screened out, providing a more accurate supplementary basis for address supplementation, making the generated address information corresponding to the address keywords more accurate and complete, and improving the accuracy of address supplementation.

[0053] In this optional implementation, based on the first task description information contained in the preset prompt word, the first model can be driven to supplement the address keyword based on the search results to generate address information corresponding to the address keyword. For example, the first task description information contained in the preset prompt word can be used to drive the first model to perform natural language understanding on the page information contained in the search results to determine the address description statement that matches the address keyword in the page information. Among them, the search results may contain at least one page information, and the first model is a model that can understand and parse the meaning and context information of natural language. The first model can perform natural language understanding on each page information to determine one or more address description statements from each page information. That is, the number of address description statements that match the address keyword determined from the page information contained in the search results can be one or more.

[0054] Based on the determined address description statement, the first model can be driven to continue to extract address names from the determined address description statement, and perform deduplication operations on the extracted address names. As mentioned above, the number of address description statements can be one or more, and the first model can extract at least one address name from each address description statement. It can be understood that the number of address names extracted by the first model from all address description statements can be one or more. The address name is a proper noun used to identify and locate a specific geographic location. It can be a city, street, community, house number, etc. Usually, the first model can extract multiple address names, and these address names come from different web page information, which may contain duplicate address names. Therefore, a deduplication operation can be performed on the extracted address names to screen out duplicate address names so that the retained address names are all different.

[0055] On this basis, the first model can continue to be driven to determine the administrative unit to which the deduplicated address name belongs. The administrative units include provinces, cities, counties, townships, villages, streets, etc. There is a hierarchical relationship between different administrative units. In this embodiment, the administrative units are hierarchical in order: provincial level (provinces, municipalities directly under the central government, autonomous regions, special administrative regions), prefecture level (cities, autonomous prefectures, autonomous leagues), county level (urban districts, county-level cities, counties, autonomous counties, banners, autonomous banners, special zones, forest areas), township level (streets, towns, townships, sumus, ethnic townships, ethnic sumus, county districts), village level (villages, communities), etc.

[0056] In the case where there are multiple address names corresponding to any administrative unit, the first model can be used to calculate the degree of fit between each address name and the address keyword under the administrative unit, and the multiple address names are sorted in descending order according to the calculated degree of fit for each address name, and other address names except the first address name are deleted from the sorted address names, so that the administrative unit corresponds to only one address name. Of course, an address name can also be randomly selected as the only address name corresponding to the address keyword under the administrative unit, and this embodiment does not limit the method of filtering address names.

[0057] Among them, when there are multiple address names corresponding to any administrative area, it can be indicated that the address keyword has a duplicate name in the map. By calculating the fitness between each address name and the address keyword, the address name with a higher fitness is selected as the address name used to supplement the address keyword. When calculating the fitness, the first model can extract the feature vectors of the address name and the address keyword, and calculate the vector distance between the feature vectors as the fitness between the address name and the address keyword. If the calculated fitness is exactly the same, then any address name is selected as the address name corresponding to the administrative unit.

[0058] For example, when the address keyword is "Tianjin Hebei", "provincial level: Tianjin City, county level: Hebei District" and "provincial level: Hebei Province" can be searched. At this time, the provincial administrative division unit corresponds to two address names. By calculating the adaptability between the address name and the address keyword, it can be understood that "Tianjin City" and "Hebei" in the address keyword have a geographical association relationship, while "Hebei Province" and "Tianjin" in the address keyword do not have a geographical association relationship. Therefore, "Tianjin City" has a higher adaptability than "Hebei Province". "Tianjin City" is used as the only address name corresponding to the address keyword under the provincial administrative division unit.

[0059] When each administrative unit corresponds to only one address name, the deduplicated address names can be concatenated according to the hierarchical relationship between the administrative units to generate address information corresponding to the address keyword.

[0060] In this implementation, address names matching the address keywords can be extracted from the search results, and the extracted address names can be filtered from two dimensions: address name and administrative unit, to ensure that there is no overlap in the address names finally screened out, and that the administrative units to which the address names finally screened out belong do not overlap, so that there is a one-to-one correspondence between the final determined address names and administrative units, which can avoid splicing errors when supplementing address keywords based on address names, thereby improving the accuracy of address supplementation.

[0061] In addition, in this embodiment, other implementations may be used to obtain search results corresponding to address keywords using the search module, and no further examples are given here.

[0062] In summary, in this embodiment, the preset search module is controlled to perform a network search based on the address keyword to obtain the search results, and based on the first task description information contained in the preset prompt word, the first model is driven to supplement the address keyword based on the search results to generate address information corresponding to the address keyword. By performing an online search for the address keyword through the search module, accurate relevant information can be obtained, especially when the administrative area information changes, the latest information can be obtained in a timely manner, thereby ensuring the accuracy and adaptation between the search results and the address keywords; based on the accurate search results, the first model is used to understand, extract and organize the search results, and the address information corresponding to the address keyword is output, thereby improving the efficiency and accuracy of address supplementation.

[0063] In the above or below embodiments, multiple implementations may be used to extract address keywords from text information.

[0064] In an optional implementation, based on the second task description information contained in the preset prompt word, the second model can be driven to perform word segmentation processing on the text information according to the pre-learned word segmentation knowledge to obtain at least one word segmentation; the at least one word segmentation is input into the label classifier respectively to obtain the label value corresponding to the at least one word segmentation; and the target word segmentation output based on the label value corresponding to the at least one word segmentation is used as the address keyword contained in the text information, and the label value corresponding to the target word segmentation is a true value.

[0065] The second model is a model that can extract address keywords from text information, and the first model and the second model can be the same model or different models. The second task description information can be used to guide the second model to extract address keywords from text information.

[0066] In this embodiment, the process of pre-training the second model may be: obtaining a large amount of annotated data, wherein the annotated data includes multiple text data, each text data includes a mapping relationship between a word and a label, and the label may be "address" or "non-address", and each label corresponds to a label value, wherein the label value corresponding to "address" is a true value, and the label value corresponding to "non-address" is a false value. Extract the features of each word in the text data, and based on the extracted features, use a machine learning algorithm (such as SVM, random forest, etc.) to train the label classifier in the second model, so that the second model can classify words using the label classifier.

[0067] In addition, by marking words labeled "address" and words labeled "non-address" in the text data and using the marked text data to train the second model, the second model can learn the word segmentation knowledge to correctly divide address word segmentations and sub-address word segmentations from the text data, which can effectively ensure the accuracy of the word segmentation granularity and avoid the inability to recognize address keywords contained in the text information due to unreasonable word segmentation.

[0068] Of course, other implementations may also be used to extract address keywords from text information, for example, address keywords may be extracted by regular expressions, etc., which is not limited here.

[0069] In summary, in this embodiment, the second model is pre-trained based on the special entity of the address, so that the second model can accurately identify the address entity from the text information, avoid dividing the address keyword into multiple participles, and destroy the semantics of the address keyword, thereby improving the accuracy of the participles, providing an accurate address basis for subsequent address supplementation, and further improving the accuracy of address supplementation.

[0070] In the above or below embodiments, the address keywords may also be supplemented by using a preset knowledge base.

[0071] In this embodiment, after the address keyword is extracted from the text information, the target address name matching the address keyword may be searched from a preset knowledge base. Figure 2 A schematic diagram of the structure of a preset knowledge base provided for an exemplary embodiment of the present application is shown in FIG. Figure 2 As shown, the preset knowledge base can be a structured knowledge base, which stores and organizes address names and the administrative units to which the address names belong in a structured manner. For example, the hierarchical relationship between address names and the administrative units to which the address names belong can be managed in the form of a tree structure, and there is a mapping relationship between address names and administrative units.

[0072] If a target address name matching the address keyword is found from the knowledge base, a candidate address name associated with the target address name can be found from the knowledge base. It can be understood that the address name on the same tree branch as the address keyword can be used as a candidate address name. Based on the mapping relationship between the address name and the administrative division unit contained in the knowledge base, the administrative division unit to which the target address name and the candidate address name belong is determined. Based on the hierarchical relationship between the administrative division units, a supplementary address name is selected from the candidate address names, and the level of the administrative division unit to which the supplementary address name belongs is higher than the level of the administrative division unit to which the target address name belongs.

[0073] It should be noted that, according to the hierarchical relationship between administrative divisions, there is usually only one high-level administrative division corresponding to a low-level administrative division, while there can be multiple low-level administrative divisions corresponding to a high-level administrative division. For example, "Chaoyang District" only belongs to "Beijing City", and there can be multiple streets under "Chaoyang District". Therefore, based on the address keyword, usually only the level of the administrative division unit higher than the address keyword can be supplemented, and the level of the administrative division unit lower than the address keyword cannot be supplemented. This is reflected in the tree structure, where the upper node (parent node, root node) of the target address name is used as the supplementary address name, and the lower nodes (child nodes, grandchild nodes) of the target address name are ignored.

[0074] On this basis, the target address names and supplementary address names can be sorted according to the hierarchical relationship between administrative units, and the sorted target address names and supplementary address names can be concatenated to generate address information corresponding to the address keywords.

[0075] Continue to refer Figure 2 , if the address keyword is "Sanlitun", the target address name "Sanlitun Street" matching the address keyword and the candidate address names "Beijing City", "Chaoyang District", "Dongsanli Community", and "Beisanli Community" associated with the target address name can be found from the knowledge base. Based on the mapping relationship between the address name and the administrative division unit, it can be determined that the administrative division unit to which "Sanlitun Street" belongs is the township level, the administrative division unit to which "Beijing City" belongs is the provincial level, the administrative division unit to which "Chaoyang District" belongs is the county level, the administrative division unit to which "Dongsanli Community" belongs is the village level, and the administrative division unit to which "Beisanli Community" belongs is the village level. Based on the hierarchical relationship between administrative division units, the candidate address names whose administrative division units have a higher level than the level of the administrative division unit to which the target address name belongs are used as supplementary address names, that is, the target address name includes "Beijing City" and "Chaoyang District". According to the hierarchical order between administrative division units, the target address name and the supplementary address name are sorted and spliced, and the address information corresponding to the address keyword can be generated as "Sanlitun Street, Chaoyang District, Beijing City".

[0076] Among them, if two identical target version names are found from the knowledge base, the adaptability between each target address name and the address keyword can be calculated, and the target address name with higher adaptability can be used as the target address name matching the address keyword, and the other target address names can be ignored; or, one of the target address names can be selected as the target address name matching the address keyword, and the other target address names can be ignored.

[0077] If the target address name matching the address keyword cannot be found from the knowledge base, the preset search module can be controlled to perform a network search based on the address keyword. The process of using the search module to perform a network search can refer to the above description and will not be repeated here.

[0078] In the case where the target address name matching the address keyword cannot be found in the knowledge base, the address information generated by the first model based on the search results can be split to obtain multiple mutually related address names and the administrative division units to which the multiple address names belong. According to the administrative division units to which the multiple address names belong, the split multiple address names are added to the preset knowledge base to update the knowledge base. In this way, the address supplementation capability of the knowledge base can be gradually improved, the frequency of use of the search module can be reduced, and thus the cost of address supplementation can be reduced.

[0079] In summary, in this embodiment, after extracting the address keywords from the text information, the address keywords are first supplemented by using the knowledge base. The knowledge base stores the address names in a structured form, which is convenient for the address supplement device to quickly locate the address names that match the address keywords, thereby improving the search efficiency. In addition, the developers can flexibly control the content in the knowledge base, and can add the flower names and common names of the region to the knowledge base to expand the knowledge base's coverage of the address keywords, thereby increasing the probability of finding the target address name therefrom. The knowledge base can also be executed offline, and address supplementation can be performed even when the network is not connected or the network status is poor. The knowledge base shares part of the work of the search module. When the knowledge base cannot achieve address supplementation, the search module is used to supplement the address, which can effectively improve the efficiency of address supplementation.

[0080] Figure 3 The following is a schematic diagram of a process in an application scenario provided by an exemplary embodiment of the present application. Figure 3 As shown, the application scenario is to automatically supplement the service address input by the user on the housekeeping platform to generate a detailed service address. In this application scenario, the address supplement device can be deployed on the housekeeping platform, and the address supplement device can be adapted to various service platforms in a pluggable form. The address supplement device includes an intelligent agent, an intelligent agent plug-in and an intelligent agent knowledge base. The intelligent agent can call the intelligent agent plug-in through an API, and the intelligent agent is mounted with an intelligent agent knowledge base.

[0081] Among them, an agent refers to an agent that can perceive the environment and take actions to achieve specific goals. It can be software, hardware or a system with autonomy, adaptability and interaction capabilities. The agent perceives changes in the environment (such as through sensors or data input), makes judgments and decisions based on the knowledge and algorithms it has learned, and then performs actions to affect the environment or achieve predetermined goals. Agent plug-ins are a set of tools that can be integrated into the agent. They allow the agent to expand its capabilities by calling external APIs, enabling it to perform more and more complex tasks. The agent knowledge base is an important component of the address supplement device for storing and managing address names, which enables the agent to understand and process information in specific fields.

[0082] In this embodiment, the housekeeping platform can provide housekeeping services to users, and users can choose appropriate housekeeping services to place orders according to their own needs. When placing an order for housekeeping services, users need to enter text information containing the user's address. The housekeeping platform can call the intelligent agent to supplement the user's address in the text information, obtain the user's detailed address, and dispatch the order according to the user's detailed address. The steps for the intelligent agent to supplement the address in this application scenario are as follows:

[0083] Step 301, receiving text information provided by the housekeeping platform, and extracting address keywords from the text information according to preset intelligent agent prompt words;

[0084] Step 302, based on the extracted address keyword, according to the preset agent prompt word, search the target address name matching the address keyword from the agent knowledge base; if the target address name is found, execute steps 303-305; if the target address name cannot be found, execute steps 306-307;

[0085] Step 303, searching the knowledge base for candidate address names associated with the target address name, and determining the administrative division unit to which the target address name belongs and the administrative division unit to which the candidate address name belongs;

[0086] Step 304: Based on the hierarchical relationship between administrative divisions, a supplementary address name is selected from the candidate address names, and the level of the administrative division unit to which the supplementary address name belongs is higher than the level of the administrative division unit to which the target address name belongs;

[0087] Step 305, sorting and concatenating the target address name and the supplementary address name according to the hierarchical relationship between the administrative division units, so as to generate address information corresponding to the address keyword based on the target address name and the supplementary address name;

[0088] Step 306, according to the preset agent prompt word, calling the agent plug-in to search the network for the address keyword to obtain the search result of the address keyword;

[0089] Step 307, supplement the address keywords based on the search results to generate address information corresponding to the address keywords.

[0090] Step 308, according to the generated address information, construct an address supplement result for the text information, and return the address supplement result to the home service platform for the home service platform to dispatch orders for users according to the address supplement result.

[0091] In this embodiment, the agent prompt is a form of input used in the field of artificial intelligence, especially in natural language processing (NLP). It is used to guide or stimulate the agent to generate specific outputs. An exemplary agent prompt provided in this embodiment is as follows:

[0092] {You are a tool that obtains provincial, municipal, district, and county information based on the input address. The obtaining steps are as follows:

[0093] 1. First, extract address-related information based on the user input. For example: input "I am in Qiqihar, Heilongjiang", extract "Qiqihar, Heilongjiang"

[0094] 2. Based on the extracted address information, obtain the complete provincial, municipal, district, and county information from the agent knowledge base in a matching manner. If no match is found, directly proceed to the next step. For example: obtain the hierarchical configuration of "Qiqihar, Heilongjiang" through matching to get Qiqihar City, Heilongjiang Province

[0095] 3. Obtain the latest information through the network search plugin and extract the complete provincial, municipal, district, and county information.

[0096] Limitations:

[0097] You only need to return in JSON format without any other additional information. The format content is as follows:

[0098] {"province": "province", "city": "city", "county": "county", "town": "town",}}.

[0099] In this embodiment, an intelligent agent supporting address supplementation is deployed on the housekeeping platform, which can provide the housekeeping platform with the ability to supplement addresses. The intelligent agent can perform a network search based on the address keywords in the text information to obtain search results containing the address keywords, and based on the natural language understanding ability, supplement the address keywords based on the search results to generate address information corresponding to the address keywords. This can avoid the situation where a complete detailed address cannot be supplemented for the address keywords due to the incomplete sub-addresses contained in the knowledge base, thereby improving the accuracy of address supplementation. Based on this, according to the generated address information, an address supplementation result is constructed for the text information, and the address supplementation result is provided to the housekeeping platform, so that the housekeeping platform can execute the business functions required by the user according to the address supplementation result, which can improve the fluency and accuracy of the housekeeping platform in executing business functions.

[0100] It should be noted that the execution subject of each step of the method provided in the above embodiment can be the same device, or the method can be executed by different devices. For example, the execution subject of steps 101 to 103 can be device A; for another example, the execution subject of steps 101 and 102 can be device A, and the execution subject of step 103 can be device B; and so on.

[0101] In addition, in some of the processes described in the above embodiments and the accompanying drawings, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this article or executed in parallel, and the sequence numbers of the operations, such as 101, 102, etc., are only used to distinguish between different operations, and the sequence numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this article are used to distinguish different messages, devices, modules, etc., do not represent the order of precedence, and do not limit the "first" and "second" to be different types.

[0102] Figure 4 FIG. 1 is a schematic diagram of a structure of an address supplement device provided by another exemplary embodiment of the present application. Figure 4 As shown, the address supplementing device may include an extraction module 41, a search module 42 and a supplementing module 43;

[0103] An extraction module 41 is used to obtain text information input by a user and perform an address extraction operation on the text information to extract address keywords from the text information;

[0104] A search module 42, configured to perform a network search based on the address keyword to obtain search results containing the address keyword, and provide the search results to the first model;

[0105] The supplementation module 43 is used to drive the first model to supplement the address keyword based on the search results based on the first task description information contained in the preset prompt word to generate address information corresponding to the address keyword; and construct an address supplementation result for the text information based on the address information generated by the first model.

[0106] In an optional embodiment, in the process of performing a network search based on the address keyword to obtain search results, the search module 42 is further used to:

[0107] Building a query request based on the address keyword, where the query request contains the address keyword;

[0108] sending a query request to a search engine associated with the search module;

[0109] The page information returned by the search engine in response to the query request is filtered to delete the page information that does not contain the address keyword, and the remaining page information after filtering is used as the search result.

[0110] In an optional embodiment, the supplementing module 43 is further used to:

[0111] The first model is driven by the first task description information contained in the preset prompt word to perform natural language understanding on the page information contained in the search results, so as to determine the address description sentence matching the address keyword in the page information, extract the address name from the determined address description sentence, and perform a deduplication operation on the extracted address name; wherein the number of the address description sentence is one or more;

[0112] Using the first model, determine the administrative units to which the duplicated address names belong; and, when each administrative unit corresponds to an address name, concatenate the duplicated address names according to the hierarchical relationship between the administrative units to generate address information corresponding to the address keyword.

[0113] In an optional embodiment, the supplement module 43 is further configured to:

[0114] When there are multiple address names corresponding to any administrative unit, the first model is used to calculate the adaptability between each address name and the address keyword under the administrative unit, and the multiple address names are sorted in order from high to low according to the adaptability calculated for each address name; and, from the sorted address names, other address names except the first address name are deleted so that the administrative unit corresponds to one address name.

[0115] In an optional embodiment, in the process of performing an address extraction operation on the text information to extract a keyword address from the text information, the extraction module 41 is further used to:

[0116] Based on the second task description information contained in the preset prompt word, drive the second model to perform word segmentation processing on the text information according to the pre-learned word segmentation knowledge to obtain at least one word segmentation; input the at least one word segmentation into the label classifier respectively to obtain the label value corresponding to the at least one word segmentation; and use the target word segmentation output based on the label value corresponding to the at least one word segmentation as the address keyword contained in the text information, and the label value corresponding to the target word segmentation is a true value.

[0117] In an optional embodiment, the address supplementing device may further include a search module, and the search module is used to:

[0118] Search the target address name that matches the address keyword from the preset knowledge base;

[0119] If the target address name matching the address keyword cannot be found from the knowledge base, the search module will perform a network search based on the address keyword.

[0120] In an optional embodiment, the search module may also be used to:

[0121] If a target address name matching the address keyword is found from the knowledge base, a candidate address name associated with the target address name is found from the knowledge base;

[0122] Based on the mapping relationship between the address name and the administrative division unit contained in the knowledge base, determining the administrative division unit to which the target address name and the candidate address name belong respectively;

[0123] Based on the hierarchical relationship between administrative divisions, a supplementary address name is selected from the candidate address names, and the level of the administrative division unit to which the supplementary address name belongs is higher than the level of the administrative division unit to which the target address name belongs;

[0124] Sort the target address names and supplementary address names according to the hierarchical relationship between administrative divisions;

[0125] The sorted target address name and the supplementary address name are concatenated to generate address information corresponding to the address keyword.

[0126] In an optional embodiment, the supplement module 43 may also be used to:

[0127] The address information generated by the first model based on the search results is split to obtain a plurality of mutually related address names and the administrative division units to which the plurality of address names respectively belong;

[0128] According to the administrative divisions to which the multiple address names respectively belong, the multiple address names that are split out are added to the knowledge base to update the knowledge base.

[0129] It is worth noting that the technical details in the above-mentioned embodiments of the address supplement device can refer to the relevant description of the action of the address supplement device in the above-mentioned embodiments of the address supplement method. In order to save space, they will not be repeated here, but this should not cause a loss in the protection scope of this application.

[0130] Figure 5 A schematic diagram of a computing device provided as another exemplary embodiment of the present application is shown in FIG. Figure 5 As shown, the computing device includes: a memory 50 , a processor 51 and a communication component 52 .

[0131] The processor 51 is coupled to the memory 50 and is used to execute the computer program in the memory 50 to:

[0132] Acquiring text information input by the user through the communication component 52;

[0133] performing an address extraction operation on the text information to extract address keywords from the text information;

[0134] Controlling a preset search module to perform a network search based on the address keyword to obtain search results containing the address keyword;

[0135] Controlling the search module to provide the search results to the first model;

[0136] Based on the first task description information contained in the preset prompt word, driving the first model to supplement the address keyword based on the search result to generate address information corresponding to the address keyword;

[0137] According to the address information generated by the first model, an address supplement result is constructed for the text information.

[0138] In an optional embodiment, in the process of controlling the preset search module to perform a network search according to the address keyword to obtain search results, the processor 51 is further configured to:

[0139] Control the search module to construct a query request based on the address keyword, where the query request contains the address keyword;

[0140] Sending the query request to a search engine associated with the search module through the search module;

[0141] The search module is used to filter the page information returned by the search engine in response to the query request to delete the page information that does not contain the address keyword, and the remaining page information after filtering is used as the search result.

[0142] In an optional embodiment, based on the first task description information included in the preset prompt word, the processor 51 is further used to drive the first model to supplement the address keyword based on the search results to generate address information corresponding to the address keyword:

[0143] The first model is driven by the first task description information contained in the preset prompt word to perform natural language understanding on the page information contained in the search results, so as to determine the address description sentence matching the address keyword in the page information, extract the address name from the determined address description sentence, and perform a deduplication operation on the extracted address name; wherein the number of the address description sentence is one or more;

[0144] Using the first model, determine the administrative units to which the duplicated address names belong; and, when each administrative unit corresponds to an address name, concatenate the duplicated address names according to the hierarchical relationship between the administrative units to generate address information corresponding to the address keyword.

[0145] In an optional embodiment, the processor 51 is further configured to:

[0146] When there are multiple address names corresponding to any administrative unit, the first model is used to calculate the adaptability between each address name and the address keyword under the administrative unit, and the multiple address names are sorted in order from high to low according to the adaptability calculated for each address name; and, from the sorted address names, other address names except the first address name are deleted so that the administrative unit corresponds to one address name.

[0147] In an optional embodiment, in the process of performing an address extraction operation on the text information to extract a keyword address from the text information, the processor 51 is further configured to:

[0148] Based on the second task description information contained in the preset prompt word, drive the second model to perform word segmentation processing on the text information according to the pre-learned word segmentation knowledge to obtain at least one word segmentation; input the at least one word segmentation into the label classifier respectively to obtain the label value corresponding to the at least one word segmentation; and use the target word segmentation output based on the label value corresponding to the at least one word segmentation as the address keyword contained in the text information, and the label value corresponding to the target word segmentation is a true value.

[0149] In an optional embodiment, in the process of controlling the preset search module to perform a network search according to the address keyword, the processor 51 is further configured to:

[0150] Search the target address name that matches the address keyword from the preset knowledge base;

[0151] If the target address name matching the address keyword cannot be found from the knowledge base, the preset search module is controlled to perform a network search based on the address keyword.

[0152] In an optional embodiment, the processor 51 is further configured to:

[0153] If a target address name matching the address keyword is found from the knowledge base, a candidate address name associated with the target address name is found from the knowledge base;

[0154] Based on the mapping relationship between the address name and the administrative division unit contained in the knowledge base, determining the administrative division unit to which the target address name and the candidate address name belong respectively;

[0155] Based on the hierarchical relationship between administrative divisions, a supplementary address name is selected from the candidate address names, and the level of the administrative division unit to which the supplementary address name belongs is higher than the level of the administrative division unit to which the target address name belongs;

[0156] Sort the target address names and supplementary address names according to the hierarchical relationship between administrative divisions;

[0157] The sorted target address name and the supplementary address name are concatenated to generate address information corresponding to the address keyword.

[0158] In an optional embodiment, the processor 51 is further configured to:

[0159] The address information generated by the first model based on the search results is split to obtain a plurality of mutually related address names and the administrative division units to which the plurality of address names respectively belong;

[0160] According to the administrative divisions to which the multiple address names respectively belong, the multiple address names that are split out are added to the knowledge base to update the knowledge base.

[0161] Further, if Figure 5 As shown, the computing device also includes: a display 53, a power supply component 54 and other components. Figure 5 Only some components are shown schematically, and it does not mean that the computing device only includes Figure 5 Components shown.

[0162] It is worth noting that the technical details in the above-mentioned embodiments of the computing device can refer to the relevant description of the actions of the computing device in the above-mentioned embodiments of the address supplement method. In order to save space, they will not be repeated here, but this should not cause a loss in the scope of protection of this application.

[0163] Accordingly, an embodiment of the present application further provides a computer-readable storage medium storing a computer program, which, when executed, can implement the steps that can be executed by a computing device in the above method embodiment.

[0164] Correspondingly, an embodiment of the present application further provides a computer program product, and when the computer program contained in the computer program is executed, the steps in the aforementioned embodiments of the address supplement method can be executed.

[0165] Above Figure 5 The memory in the computer is used to store computer programs and can be configured to store various other data to support operations on the computing platform. Examples of such data include instructions for any application or method operating on the computing platform, contact data, phone book data, messages, pictures, videos, etc. The memory can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0166] Above Figure 5 The communication component in is configured to facilitate wired or wireless communication between the device where the communication component is located and other devices. The device where the communication component is located can access a wireless network based on a communication standard, such as WiFi, 2G, 3G, 4G / LTE, 5G and other mobile communication networks, or a combination thereof. In an exemplary embodiment, the communication component receives a broadcast signal or broadcast-related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component also includes a near field communication (NFC) module to facilitate short-range communication. For example, the NFC module can be implemented based on radio frequency identification (RFID) technology, infrared data association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology and other technologies.

[0167] Above Figure 5 The display in the embodiment includes a screen, which may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive input signals from a user. The touch panel includes one or more touch sensors to sense touch, slide, and gestures on the touch panel. The touch sensor may not only sense the boundary of a touch or slide action, but also detect the duration and pressure associated with the touch or slide operation.

[0168] Above Figure 5 The power supply component in the device provides power to various components of the device where the power supply component is located. The power supply component may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power to the device where the power supply component is located.

[0169] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0170] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0171] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0172] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0173] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0174] The memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0175] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0176] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0177] It should be noted that 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, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0178] The above is only an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. An address supplement method, characterized in that: Applicable to an address supplement device, the method comprises: Get the text information entered by the user; performing an address extraction operation on the text information to extract address keywords from the text information; Controlling a preset search module to perform a network search according to the address keyword to obtain search results containing the address keyword; Controlling the search module to provide the search results to the first model; Based on the first task description information included in the preset prompt word, driving the first model to supplement the address keyword based on the search result to generate address information corresponding to the address keyword; An address supplement result is constructed for the text information according to the address information generated by the first model.

2. The method according to claim 1, characterized in that Controlling a preset search module to perform a network search based on the address keyword to obtain search results, including: Controlling the search module to construct a query request based on the address keyword, wherein the query request includes the address keyword; Sending the query request to a search engine associated with the search module through the search module; Using the search module, filtering the page information returned by the search engine in response to the query request to delete the page information that does not contain the address keyword; The remaining page information after filtering is used as the search result.

3. The method according to claim 1, characterized in that Based on the first task description information included in the preset prompt word, driving the first model to supplement the address keyword based on the search result to generate address information corresponding to the address keyword, including: The first model is driven by the first task description information contained in the preset prompt word to perform natural language understanding on the page information contained in the search result, so as to determine the address description sentence matching the address keyword in the page information, extract the address name from the determined address description sentence, and perform a deduplication operation on the extracted address name; wherein the number of the address description sentence is one or more; Using the first model, determine the administrative unit to which the duplicated address name belongs; and, when each administrative unit corresponds to an address name, concatenate the duplicated address names according to the hierarchical relationship between the administrative units to generate address information corresponding to the address keyword.

4. The method according to claim 3, characterized in that Also includes: In the case where there are multiple address names corresponding to any administrative unit, the first model is used to calculate the degree of fit between each address name under the administrative unit and the address keyword, and the multiple address names are sorted in descending order according to the calculated degree of fit for each address name; And, from the sorted address names, other address names except the first address name are deleted so that the administrative unit corresponds to one address name.

5. The method according to claim 1, characterized in that Performing an address extraction operation on the text information to extract a keyword address from the text information includes: Based on the second task description information contained in the preset prompt word, drive the second model to perform word segmentation processing on the text information according to the pre-learned word segmentation knowledge to obtain at least one word segmentation; input the at least one word segmentation into the label classifier respectively to obtain the label value corresponding to each of the at least one word segmentation; and use the target word segmentation output based on the label value corresponding to each of the at least one word segmentation as the address keyword contained in the text information, and the label value corresponding to the target word segmentation is a true value.

6. The method according to claim 1, characterized in that Controlling a preset search module to perform a network search based on the address keyword also includes: Searching for a target address name that matches the address keyword from a preset knowledge base; If the target address name matching the address keyword cannot be found from the knowledge base, the preset search module is controlled to perform a network search according to the address keyword.

7. The method according to claim 6, characterized in that Also includes: If a target address name matching the address keyword is found from the knowledge base, then a candidate address name associated with the target address name is found from the knowledge base; Based on the mapping relationship between the address name and the administrative division unit contained in the knowledge base, determining the administrative division unit to which the target address name and the candidate address name respectively belong; Based on the hierarchical relationship between administrative divisions, selecting a supplementary address name from the candidate address names, the level of the administrative division unit to which the supplementary address name belongs is higher than the level of the administrative division unit to which the target address name belongs; Sorting the target address name and the supplementary address name according to the hierarchical relationship between administrative divisions; The sorted target address name and the supplementary address name are concatenated to generate address information corresponding to the address keyword.

8. The method according to claim 6, characterized in that Also includes: Splitting the address information generated by the first model based on the search results to obtain a plurality of mutually related address names and the administrative division units to which the plurality of address names respectively belong; According to the administrative divisions to which the multiple address names respectively belong, the multiple address names separated out are added to the knowledge base to update the knowledge base.

9. An address supplement device, characterized in that: The address supplement device comprises an extraction module, a search module and a supplement module; The extraction module is used to obtain text information input by the user, and perform an address extraction operation on the text information to extract address keywords from the text information; The search module is used to perform a network search based on the address keyword to obtain search results containing the address keyword, and provide the search results to the supplement module; The supplementing module is used to drive the first model to supplement the address keyword based on the search results based on the first task description information contained in the preset prompt word to generate address information corresponding to the address keyword; and construct an address supplement result for the text information based on the address information generated by the first model.

10. A computing device, characterized in that: Includes memory, processor, and communication components; The memory is used to store one or more computer instructions; The processor is coupled to the memory and the communication component, and is configured to execute the one or more computer instructions to execute the address supplement method according to any one of claims 1 to 8.

11. A computer-readable storage medium storing computer instructions, characterized in that: When the computer instructions are executed by one or more processors, the one or more processors are caused to execute the address supplement method according to any one of claims 1 to 8.

Images