Business processing method based on hierarchical disambiguation

Abstract

The application provides a business processing method based on hierarchical disambiguation, which can be applied to the technical field of computers. The method comprises the following steps: in response to receiving a business processing request, performing semantic analysis on the business processing request, determining a geographical slot dictionary and a business type; in the case that the business type is a target business type, performing division suffix stripping according to the longest first principle to obtain a target business processing request and a target standard administrative division text; in response to the existence of a target value in a region identifier field in the geographical slot dictionary, querying the target value from standard region comparison information to obtain a region query result; in response to the region query result representing the existence of a target region and a jurisdictional region, determining a disambiguation level based on matching information between the target business processing request and the target standard administrative division text; and in response to the disambiguation level being a first level, calling a service interface corresponding to the business type to process the business processing request to generate a business processing result.

Description

Technical Field

[0001] This application relates to the field of computer technology, and more specifically to a business processing method based on hierarchical disambiguation. Background Technology

[0002] In a vertically oriented conversational query system, users' natural language input typically includes geographic location entities. The system needs to map these natural language entity representations to specific geographic location entities in the backend standard database in order to correctly invoke downstream service interfaces to obtain query results.

[0003] In related technologies, the above mapping is usually implemented using a Large Language Model (LLM) or keyword matching. However, the judgment of LLM is random and uncontrollable, which can easily lead to errors or failures in downstream service interface calls; keyword matching is difficult to effectively handle non-standard name variations, which can also easily cause mapping errors, resulting in abnormal downstream service interface calls. Summary of the Invention

[0004] In view of the above problems, this application provides a business processing method based on hierarchical disambiguation.

[0005] According to one aspect of this application, a hierarchical disambiguation business processing method is provided, comprising: in response to receiving a business processing request, performing semantic parsing on the business processing request to determine the geographic slot dictionary and business type corresponding to the business processing request; if the business type is a target business type, according to the longest priority principle, stripping the administrative division text included in the business processing request and the standard administrative division text corresponding to the administrative division text in the standard administrative division library by removing the administrative division suffix, to obtain the target business processing request and the target standard administrative division text; in response to the existence of a target value in the regional identifier field of the geographic slot dictionary, querying the target region corresponding to the target value and the jurisdiction to which the target region belongs from the standard regional comparison information to obtain a regional query result; in response to the regional query result indicating the existence of a target region and a jurisdiction, determining the disambiguation level of the target region based on the matching information between the target business processing request and the target standard administrative division text; in response to the disambiguation level being the first level, calling the service interface corresponding to the business type to process the business processing request, so that the service interface generates a business processing result corresponding to the target region.

[0006] Another aspect of this application provides a business processing apparatus based on hierarchical disambiguation, comprising: a parsing module, configured to, in response to receiving a business processing request, perform semantic parsing on the business processing request to determine the geographic slot dictionary and business type corresponding to the business processing request; a stripping module, configured to, when the business type is a target business type, strip the administrative division text included in the business processing request and the standard administrative division text corresponding to the administrative division text in the standard administrative division library according to the longest priority principle, to obtain the target business processing request and the target standard administrative division text; a query module, configured to, in response to the existence of a target value in the regional identifier field of the geographic slot dictionary, query the target region corresponding to the target value and the jurisdiction to which the target region belongs from the standard regional comparison information to obtain a regional query result; a determination module, configured to, in response to the regional query result indicating the existence of a target region and a jurisdiction, determine the disambiguation level of the target region based on the matching information between the target business processing request and the target standard administrative division text; and a processing module, configured to, in response to the disambiguation level being the first level, call the service interface corresponding to the business type to process the business processing request, so that the service interface generates a business processing result corresponding to the target region.

[0007] Another aspect of this application provides an electronic device comprising: one or more processors; and a memory for storing one or more computer programs, wherein the one or more processors execute the one or more computer programs to implement the steps of the method described above.

[0008] Another aspect of this application provides a computer-readable storage medium having a computer program or instructions stored thereon, which, when executed by a processor, implement the steps of the above-described method.

[0009] Another aspect of this application provides a computer program product, including a computer program or instructions that, when executed by a processor, implement the steps of the above-described method.

[0010] According to embodiments of this application, by semantically parsing business processing requests, the corresponding geographic slot dictionary and business type are determined, thereby enabling structured extraction of geographic entity information from the business processing request and triggering subsequent hierarchical disambiguation processes as needed based on the business type, reducing invalid calculations for irrelevant businesses. When the business type is determined to be the target business type, the administrative division suffixes are stripped from the business processing request and the standard administrative division text in the standard administrative division library according to the longest priority principle, to reduce interference caused by differences in administrative division suffixes on text comparison and improve the tolerance for non-standard variations in user descriptions. Next, it is checked whether the region identifier field in the geographic slot dictionary contains a target value. If it does, the target region and its jurisdiction are queried from the standard region comparison information. If the target region and jurisdiction are found, the disambiguation level of the target region is determined based on the matching information between the target business processing request and the target standard administrative division text, thereby quantifying the confidence level of the aforementioned geographic slot dictionary. Furthermore, when the disambiguation level is determined to be Level 1, meaning the geographic slot dictionary is reliable, the service interface corresponding to the business type is invoked to generate the business processing result. This ensures that the service interface call is executed only when the entity resolution is relatively reliable, reducing the generation of erroneous results and improving the accuracy of business processing. Therefore, this at least partially solves the technical problem in related technologies where the low accuracy of geographic entity mapping leads to abnormal downstream service interface calls, achieving the technical effect of improving the accuracy of geographic entity mapping and business processing. Attached Figure Description

[0011] The above-mentioned contents, other objects, features and advantages of this application will become clearer from the following description of embodiments of this application with reference to the accompanying drawings.

[0012] Figure 1 The diagram illustrates an application scenario of a hierarchical disambiguation-based service processing method and apparatus according to an embodiment of this application.

[0013] Figure 2 A flowchart of a service processing method based on hierarchical disambiguation according to an embodiment of this application is shown.

[0014] Figure 3 A flowchart illustrating the process of stripping suffixes according to an embodiment of this application is shown.

[0015] Figure 4 A flowchart illustrating the determination of the disambiguation level according to an embodiment of this application is shown.

[0016] Figure 5 An application scenario diagram of a service processing method and apparatus based on hierarchical disambiguation according to another embodiment of this application is shown.

[0017] Figure 6A structural block diagram of a service processing apparatus based on hierarchical disambiguation according to an embodiment of this application is shown.

[0018] Figure 7 A block diagram of an electronic device suitable for implementing a hierarchical disambiguation-based service processing method according to an embodiment of this application is shown. Detailed Implementation

[0019] The embodiments of this application will now be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of this application. In the following detailed description, numerous specific details are set forth to provide a thorough understanding of the embodiments of this application for ease of explanation. However, it will be apparent that one or more embodiments may be implemented without these specific details. Furthermore, descriptions of well-known structures and technologies are omitted in the following description to avoid unnecessarily obscuring the concepts of this application.

[0020] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of this application. The terms “comprising,” “including,” etc., as used herein indicate the presence of the stated features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0021] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein are to be interpreted in a manner consistent with the context of this specification, and not in an idealized or overly rigid way.

[0022] When using expressions such as "at least one of A, B and C", they should generally be interpreted in accordance with the meaning that is commonly understood by those skilled in the art (e.g., "a system having at least one of A, B and C" should include, but is not limited to, a system having A alone, a system having B alone, a system having C alone, a system having A and B, a system having A and C, a system having B and C, and / or a system having A, B and C, etc.).

[0023] During the research, it was found that for a pure LLM free dialogue scheme, the LLM's judgment is random and uncontrollable; multiple executions of the same input may yield different conclusions. The LLM has a weak ability to accurately perceive standard database entries, and is prone to mismatching or missing non-standard names (such as colloquial place names, abbreviations of place names, etc.) to incorrect entries.

[0024] Pure rule-based or keyword matching schemes, which use traditional exact string matching or keyword matching (such as regular expressions) to directly map user input to standard library entries, struggle to handle non-standard name variations. For example, "[Province A, City B]" and "[Province A, City B]" refer to the same city, but exact matching will fail. Furthermore, they struggle to handle the diversity of administrative division suffixes, such as "Autonomous Region A," "District A," and "A," which may be used by users to refer to the same geographical location in actual conversations, leading to mapping failures or errors.

[0025] Furthermore, in related technologies, the matching results are usually only binary results of "[hit]" or "[miss]", making it difficult for the system to provide user-friendly guidance for situations that are "close but not a perfect match".

[0026] In view of this, embodiments of this application provide a business processing method based on hierarchical disambiguation, comprising: in response to receiving a business processing request, performing semantic parsing on the business processing request to determine a geographic slot dictionary and a business type corresponding to the business processing request; if the business type is a target business type, according to the longest priority principle, stripping the administrative division text included in the business processing request and the standard division text corresponding to the administrative division text in the standard division library by removing the division suffix, to obtain a target business processing request and a target standard administrative division text; in response to the existence of a target value in the regional identifier field of the geographic slot dictionary, querying the target region corresponding to the target value and the jurisdiction to which the target region belongs from the standard region comparison information to obtain a region query result; in response to the region query result indicating the existence of the target region and the jurisdiction, determining the disambiguation level of the target region based on the matching information between the target business processing request and the target standard administrative division text; in response to the disambiguation level being the first level, calling a service interface corresponding to the business type to process the business processing request, so that the service interface generates a business processing result corresponding to the target region.

[0027] Figure 1 The diagram illustrates an application scenario of a hierarchical disambiguation-based service processing method and apparatus according to an embodiment of this application.

[0028] like Figure 1 As shown, application scenario 100 according to this embodiment may include a first terminal device 101, a second terminal device 102, a third terminal device 103, a network 104, and a server 105. The network 104 serves as a medium for providing communication links between the first terminal device 101, the second terminal device 102, the third terminal device 103, and the server 105. The network 104 may include various connection types, such as wired or wireless communication links, or fiber optic cables, etc.

[0029] Users can use the first terminal device 101, the second terminal device 102, and the third terminal device 103 to interact with the server 105 via the network 104 to receive or send messages, etc. Various communication client applications can be installed on the first terminal device 101, the second terminal device 102, and the third terminal device 103, such as shopping applications, web browser applications, search applications, instant messaging tools, email clients, social media platform software, etc. (for example only).

[0030] The first terminal device 101, the second terminal device 102, and the third terminal device 103 can be various electronic devices with displays and support web browsing, including but not limited to smartphones, tablets, laptops, and desktop computers.

[0031] Server 105 can be a server that provides various services, such as a backend management server that supports business processing requests sent by users using the first terminal device 101, the second terminal device 102, and the third terminal device 103 (this is just an example). The backend management server can analyze and process the received business processing requests and other data, and feed back the processing results (such as business processing results corresponding to the target area) to the terminal devices.

[0032] It should be noted that the hierarchical disambiguation-based service processing method provided in this application embodiment can generally be executed by server 105. Correspondingly, the hierarchical disambiguation-based service processing device provided in this application embodiment can generally be located in server 105. The hierarchical disambiguation-based service processing method provided in this application embodiment can also be executed by a server or server cluster that is different from server 105 and capable of communicating with the first terminal device 101, the second terminal device 102, the third terminal device 103, and / or server 105. Correspondingly, the hierarchical disambiguation-based service processing device provided in this application embodiment can also be located in a server or server cluster that is different from server 105 and capable of communicating with the first terminal device 101, the second terminal device 102, the third terminal device 103, and / or server 105.

[0033] It should be understood that Figure 1 The number of first terminal devices, second terminal devices, third terminal devices, networks, and servers shown in the diagram is merely illustrative. Depending on implementation needs, any number of first terminal devices, second terminal devices, third terminal devices, networks, and servers can be included.

[0034] The following will be based on Figure 1 The described scene, through Figures 2-6 The application embodiment provides a detailed description of the hierarchical disambiguation-based service processing method.

[0035] Figure 2 A flowchart of a service processing method based on hierarchical disambiguation according to an embodiment of this application is shown.

[0036] like Figure 2 As shown, the method includes operations S210 to S250.

[0037] In operation S210, in response to receiving a business processing request, semantic parsing is performed on the business processing request to determine the geoslot dictionary and business type corresponding to the business processing request.

[0038] In operation S220, when the business type is the target business type, the administrative division text included in the business processing request and the standard administrative division text corresponding to the administrative division text in the standard administrative division library are stripped of their administrative division suffixes according to the longest priority principle, so as to obtain the target business processing request and the target standard administrative division text.

[0039] In operation S230, in response to the existence of a target value in the area identifier field of the geographic slot dictionary, the target area corresponding to the target value and the jurisdiction to which the target area belongs are queried from the standard area comparison information to obtain the area query results.

[0040] In operation S240, in response to the regional query results indicating the existence of a target region and a jurisdiction, the disambiguation level of the target region is determined based on the matching information between the target business processing request and the target standard administrative division text.

[0041] When operating S250, in response to the disambiguation level being the first level, the service interface corresponding to the business type is called to process the business processing request, so that the service interface generates a business processing result corresponding to the target region.

[0042] In implementation, business processing requests can originate from a target object, which can be a user or a client device operated by the user. The business processing request can be input text from the target object, including administrative division text (i.e., textual descriptions corresponding to geographical location entities) and descriptions related to the business type. Alternatively, in some embodiments, the business processing request may only include descriptions related to the business type. Therefore, semantic parsing of the business processing request is necessary to determine whether it contains administrative division text and to identify the business indicated by the request.

[0043] There are no restrictions on the method of semantic parsing the business processing request. For example, in one embodiment of this application, the semantic parsing interface of the Large Language Model (LLM) can be called to perform structured parsing of the request text and obtain the corresponding geographic slot dictionary and the business type to which this request belongs (e.g., annual energy consumption analysis, energy consumption ranking, compliance check, local standard query, energy consumption assessment report, etc.).

[0044] If the current business type is determined to be a target business type requiring hierarchical disambiguation, a hierarchical disambiguation process can be triggered. For example, suffix stripping can be performed on business processing requests and the standard administrative division library according to the longest priority principle. This longest priority principle means that when stripping administrative division suffixes from the administrative division text in the business processing request and the corresponding standard administrative division text in the standard administrative division library, the longest administrative division suffix is ​​stripped first, followed by the second longest, and so on, until all strippable suffixes are stripped.

[0045] The standard administrative division database can be stored in the form of a lookup table. Each record contains the following fields: region identifier, jurisdictional region name, and target region name. The region identifier is a unique identifier (e.g., cityId) used to uniquely represent the corresponding administrative division; it can be stored as an integer. The jurisdictional region name is the name of the superior administrative region to which the target region belongs (which can be one or more). The target region name is the name of a queryable geographical unit (e.g., city, county, league), which can be stored as a string. For example, a record stored in the standard administrative division database could be 001A Province B City, indicating that the region identifier is 001, and the corresponding geographical entity is Province A / City B, where City B is the target region (city), and Province A is the jurisdictional region to which City B belongs.

[0046] Figure 3 A flowchart illustrating the process of stripping suffixes according to an embodiment of this application is shown.

[0047] like Figure 3 As shown, taking a business processing request containing multiple levels of administrative divisions as an example, suppose the business processing request includes administrative division text of "Autonomous Region A, City B". The administrative division suffixes in the business processing request can be stripped according to the suffix stripping list determined by the longest priority principle. During the stripping process, stripping is attempted in descending order of priority (i.e., from longest to shortest suffix character length).

[0048] The specific process is as follows: First, an attempt is made to extract "X Ethnic Autonomous Region" (i.e., the complete suffix pattern containing the ethnic name prefix). Since the input text "A Autonomous Region B City" does not contain this complete pattern (e.g., "X Ethnicity" and "A" are inconsistent), this attempt fails. The next attempt is made to extract the pattern "Y Ethnic Autonomous Region" (corresponding to another ethnic name prefix), which also fails. The next attempt is made to extract the pattern "X Ethnic Autonomous Region," which also fails. The next attempt is made to extract the general suffix "Autonomous Region," which is found in the input text "A Autonomous Region B City," resulting in the intermediate result "AB City." Subsequently, attempts are made to extract "Region" and "Province" in priority order, all of which fail. The next attempt is made to extract "City," which is found in the intermediate result "AB City," resulting in the final result "AB." Finally, attempts are made to extract "County" and "League," all of which fail, and the extraction process ends.

[0049] It should be noted that the above administrative division suffix types and priority order are for illustrative purposes only. Other administrative division suffix types can be matched during the suffix stripping process as needed, and an adaptive priority order can be set.

[0050] The aforementioned suffix stripping list can be determined in the following way: for example, based on a full scan analysis of the standard database of place names (or standard administrative division database), the initial suffix stripping list is obtained by arranging them according to the longest priority principle. This is because there is an inclusion relationship between administrative division suffixes. For example, "A-ethnic autonomous region" may match both the complete suffix pattern (longer in length) containing the ethnic name prefix and the general suffix pattern "autonomous region" (shorter in length). If the shorter general suffix "autonomous region" is stripped first, "A-ethnic autonomous region" will be stripped to "A-ethnic," and subsequent stripping of the ethnic name prefix will no longer match, resulting in residual suffixes. Therefore, placing the complete suffix containing the ethnic name prefix before the general suffix ensures that longer suffixes are stripped first, reducing the problem of longer suffixes being left behind due to the stripping of shorter suffixes first.

[0051] Furthermore, the initial suffix stripping list above excludes predetermined administrative division suffixes (such as "state" and "district"), resulting in the final suffix stripping list. This is because if "state" were included in the stripping process, it would corrupt numerous place names containing the character "state," causing them to be incorrectly stripped into single characters and thus losing their recognizability. Similarly, if "district" were included in the stripping process alone, it would corrupt place names containing the suffix "region," causing the "district" part of "region" to be stripped prematurely, making subsequent "region" suffixes unmatchable as a whole. Therefore, excluding "state" and individual "district" suffixes from the stripping scope improves the accuracy of the stripping operation and the completeness of the stripped administrative divisions.

[0052] You can check if the region identifier field (e.g., cityId) in the geographic slot dictionary has a target value. If it does, use that target value as the query key to find the corresponding target region (e.g., city name) and the jurisdiction (e.g., province name) to which the target region belongs from the standard region comparison information (e.g., standard city database comparison table), and obtain the region query results.

[0053] Standard area comparison information may include a mapping relationship between area identifiers and areas, used to find the corresponding target area and its jurisdiction based on the area identifier. In some embodiments, the storage format of the standard area comparison information may differ from that of the standard regional division database. In other embodiments, the standard regional division database may include the aforementioned standard area comparison information.

[0054] If the corresponding target region and jurisdiction are found, the disambiguation level of the target region can be output based on the matching information between the stripped target business processing request and the target standard administrative division text.

[0055] When the disambiguation level is Level 1 (which indicates an exact match, corresponding to high confidence), the downstream service interface matching the business type is invoked, with the target region passed as a parameter. The service interface then generates and returns the corresponding business processing result. The downstream service interface can be an Application Programming Interface (API) corresponding to each business type.

[0056] According to embodiments of this application, by semantically parsing business processing requests, the corresponding geographic slot dictionary and business type are determined, thereby enabling structured extraction of geographic entity information from the business processing requests and triggering subsequent hierarchical disambiguation processes as needed based on the business type, reducing invalid calculations for irrelevant businesses. When the business type is determined to be the target business type, the administrative division suffixes are stripped from the business processing request and the standard administrative division text in the standard administrative division library according to the longest priority principle, to reduce interference caused by differences in administrative division suffixes on text comparison and improve the tolerance for non-standard variations in user descriptions. Next, it is checked whether the region identifier field in the geographic slot dictionary contains a target value. If it does, the target region and its jurisdiction are queried from the standard region comparison information. If the target region and jurisdiction are found, the disambiguation level of the target region is determined based on the matching information between the target business processing request and the target standard administrative division text, thereby quantifying the confidence level of the aforementioned geographic slot dictionary. Furthermore, when the disambiguation level is determined to be Level 1, meaning the geographic slot dictionary is reliable, the service interface corresponding to the business type is called to generate the business processing result. This ensures that the service interface call is executed only when the entity resolution is relatively reliable, reducing the generation of erroneous results and improving the accuracy of business processing. Therefore, this approach at least partially solves the technical problem in related technologies where the low accuracy of geographic entity mapping leads to abnormal downstream service interface calls, achieving the technical effect of improving both geographic entity mapping accuracy and business processing accuracy.

[0057] According to an embodiment of this application, in response to a disambiguation level of Level 1, a service interface corresponding to the business type is invoked to process the business processing request, so that the service interface generates a business processing result corresponding to the target region. This includes: in response to a disambiguation level of Level 1, generating a release signal corresponding to the business processing request and a call request for invoking the service interface corresponding to the business type, wherein the call request includes text corresponding to the target region and the jurisdiction region concatenated according to a preset format; in response to the release signal, invoking the corresponding service interface based on the call request, so that the service interface generates a business processing result corresponding to the target region.

[0058] When the disambiguation level is Level 1, a clearance signal can be generated, and a call request for invoking the corresponding service interface can be constructed. This call request includes the target region and jurisdiction region concatenated according to a preset format (e.g., formatted as "province / city"). The clearance signal can be used for gating decisions. This clearance signal can be obtained by the auto-fetch node. If the signal is identified as a clearance signal, conditional routing can send the business processing request and the call request, or only the call request, to the API call node to invoke the corresponding service interface.

[0059] This allows the service interface to execute specific business logic and return the business processing result corresponding to the target region. The above concatenation format helps downstream service interfaces to uniformly parse geographic location parameters, reducing call anomalies caused by inconsistent formats.

[0060] According to an embodiment of this application, by using a release signal, the business processing request is routed to the service interface call node corresponding to the business type, so that the business processing request enters the service interface call chain, reducing the situation where downstream service interface calls are erroneously triggered when the confidence level is not reached.

[0061] According to an embodiment of this application, semantic parsing is performed on a business processing request to determine the geographic slot dictionary and business type corresponding to the business processing request, including: performing semantic parsing on the business processing request to obtain a parsing result; if the parsing result indicates that the business processing request does not include administrative division text, determining the historical geographic slot dictionary corresponding to the target object based on the target object sending the business processing request; using the historical geographic slot dictionary as the geographic slot dictionary corresponding to the business processing request; and determining the business type corresponding to the business processing request based on the parsing result.

[0062] During implementation, the LLM interface can be called to obtain semantic parsing results. If the parsing result indicates that the current request does not contain any administrative division text (e.g., the target object only says "help me check the compliance status" without mentioning any geographical entities), it means that the target object did not actively mention geographical information in this round. In this case, instead of directly returning an empty slot, based on the target object that sent the request (e.g., determined by the dialogue session identifier), the historical geographical slot dictionary corresponding to the target object is obtained from the current dialogue context, and this historical dictionary is reused as the geographical slot dictionary for the current request. The business type is still determined based on the parsing result. This implementation method allows the historical parsing result to be automatically used even when the user omits repeated place names in multiple rounds of dialogue, reducing invalid follow-up questions.

[0063] In other embodiments, if the parsing result indicates that the business processing request includes administrative division text, it means that the target object has actively mentioned geographical information in this round. In this case, based on the administrative division text extracted from the parsing result (e.g., "Province A, City B"), the regional identifier and standard division entries corresponding to the administrative division text can be queried from the standard division database to generate a geographic slot dictionary corresponding to the current request. The standard division entries at least include the target area and the jurisdictional area.

[0064] The specific implementation process can be as follows: Extract the administrative division text mentioned by the target object from the parsing results. The administrative division text may include the name of the target area (such as "City B"), and optionally the name of the jurisdiction (such as "Province A"). Using the name of the target area and / or the name of the jurisdiction as the query key, perform a matching query in the standard administrative division database to determine the area identifier (cityId) and standard administrative division entry (such as "Province A / City B") corresponding to the administrative division text.

[0065] After retrieving the corresponding region identifier and standard zoning entry, the region identifier can be used as the target value of the region identifier field, and the target region name in the standard zoning entry can be used as the target string of the region name field to generate a geoslot dictionary corresponding to the business processing request.

[0066] If no record corresponding to the administrative division text is found in the standard zoning database, it means that the geographic entity mentioned by the target object is not within the supported scope. In this case, an empty geographic slot dictionary can be generated.

[0067] According to the embodiments of this application, the above method can locate geographic entities and generate a slot dictionary based on a standard division database when the current request contains administrative division text, and inherit historical slot information when the current request does not contain administrative division text, thereby taking into account both accurate single-turn parsing and multi-turn dialogue continuity.

[0068] According to an embodiment of this application, the target standard administrative division text includes a standard target region text corresponding to the target region and a standard jurisdictional region text corresponding to the jurisdictional region; wherein, determining the disambiguation level of the target region based on the matching information between the target business processing request and the target standard administrative division text includes: comparing the target business processing request with the target standard administrative division text to determine the presence of the standard target region text and the standard jurisdictional region text in the target business processing request; and determining the disambiguation level of the target region based on the presence of the standard target region text and the standard jurisdictional region text in the target business processing request.

[0069] The standard target area text can be the name of the target area after separation (such as "B"), and the standard jurisdiction area text can be the name of the jurisdiction area after separation (such as "A"). Together, they constitute the target standard administrative division text (such as "AB").

[0070] When determining the disambiguation level of a target region based on matching information, the stripped target business processing request (i.e., user-input compressed text) can be compared with the target standard administrative division text (i.e., standard entry compressed text). During the comparison, the presence of the standard target region text and the standard jurisdiction region text in the target business processing request can be detected separately, thereby determining their presence in the target business processing request. These presence scenarios include, but are not limited to: the standard target region text existing independently in the target business processing request; the standard jurisdiction region text existing independently in the target business processing request; both the standard target region text and the standard jurisdiction region text existing simultaneously in the target business processing request; or neither existing in the target business processing request.

[0071] Subsequently, the disambiguation level of the target area can be determined based on the aforementioned conditions. For example, when both the standard target area text and the standard jurisdiction area text exist in the target business processing request, it indicates that the target object's input in this round fully covers the target standard administrative division text, and the disambiguation level can be determined as Level 1. When neither exists in the target business processing request, it indicates that the user has not mentioned relevant information about the geographic entity in this round, and it can be further determined whether there is relevant information about the geographic entity in the business processing request, and further graded disambiguation can be performed based on the identification results.

[0072] According to embodiments of this application, through the above comparison and determination of existence, it is possible to determine whether the target object's current request contains the target area and its jurisdiction, thereby determining whether the above-mentioned geographic slot dictionary is reliable and its degree of reliability. Throughout the determination process, comparison can be performed solely based on the stripped business processing requests, without incorporating the standardized area names already stored in the slots. This reduces the likelihood of misjudging as an exact match due to the write-back standard name being identical to the database standard name, ensuring the objectivity and accuracy of the disambiguation process.

[0073] According to an embodiment of this application, determining the disambiguation level of a target region based on the presence of standard target region text and standard jurisdiction region text in a target service processing request further includes: determining the disambiguation level as a first level if the presence meets preset existence conditions; wherein the preset existence conditions include at least one of the following: the standard target region text exists in the target service processing request; and the standard target region text and the standard jurisdiction region text exist simultaneously in the target service processing request.

[0074] The preset existence conditions are used to determine the degree of matching between the target business processing request and the target standard administrative division text, thereby verifying whether the geographic entity represented by the region identifier field in the geographic slot dictionary matches the target object's current business processing request. Since both the target standard administrative division text and the geographic slot dictionary are determined through a standard division library, comparing the target standard administrative division text with the user's compressed text can verify whether the geographic slot dictionary accurately reflects the geographic intent in the target object's current request.

[0075] Based on the above principles, the preset existence conditions may include at least one of the following: the standard target area text (no less than 2 characters) exists in the user's compressed text; and the standard target area text and the standard jurisdiction area text both exist in the user's compressed text. If either of the above conditions is met, it indicates that the user's current request matches the standard administrative division entry, the area identifier field in the slot is reliable, and the disambiguation level is determined to be Level 1.

[0076] According to the embodiments of this application, the above method is used to verify the results of the extraction of the geographic slot dictionary by comparing the stripped text, thereby improving the accuracy of geographic entity mapping.

[0077] According to an embodiment of this application, determining the disambiguation level of a target region based on the presence of standard target region text and standard jurisdiction text in a target business processing request includes: when the presence does not meet preset presence conditions, performing region-related detection on the business processing request based on the presence of standard target region text and standard jurisdiction text in the target business processing request, and the presence of administrative division suffixes in the business processing request, to obtain detection results; and determining the disambiguation level as the first level when the detection results indicate that the business processing request does not involve administrative division text related to the target region.

[0078] When the above conditions are not met, it indicates that the stripped standard administrative division text was not fully or partially matched in the user's compressed text. In this case, the next round of regional involvement detection can be initiated to determine whether the business processing request includes administrative division text and whether it involves administrative division text related to the target region.

[0079] Regional involvement detection can be comprehensively determined based on the following preset detection conditions: whether the standard target region text (i.e., the compressed text of the target region name, with a length of no less than 2 characters) exists in the target business processing request; whether the standard jurisdiction region text (i.e., the compressed text of the jurisdiction region name, with a length of no less than 2 characters) exists in the target business processing request; and whether the business processing request (i.e., the user's original input text) contains administrative division suffix characters (including province, city, district, county, prefecture, league, autonomous region, etc.). Regional involvement detection can be performed on the business processing request based on the above preset detection conditions to obtain the detection results.

[0080] If the detection results show that none of the above preset detection conditions are met, meaning that neither the standard target area text nor the standard jurisdiction area text appears in the user's compressed text, and the user's original input text does not contain any administrative division suffix characters, then it indicates that the user's request in this round did not involve any administrative division text related to the target area. In this case, it can be determined that the target object did not raise any objection to the historical parsing results (e.g., historical geographic slot dictionary) in this round, and the existing regional identifiers in the geographic slot dictionary are trusted by default, and the disambiguation level is determined to be the first level.

[0081] According to embodiments of this application, in a multi-turn dialogue, a user may only express business needs in the current turn (such as "help me check the compliance status"), without repeating previously specified geographic entity information. In this case, the correctness of the parsed results already in the slot should not be questioned because geographic information was not mentioned in this turn. Instead, the confirmed geographic entity information should be used by default, thereby reducing unnecessary follow-up questions and improving the coherence of the dialogue.

[0082] The aforementioned minimum valid character length (no less than 2 characters) restriction can reduce false matches caused by single residual characters in the suffix (such as the remaining "state" or "province" after stripping), further improving the reliability of detection.

[0083] According to an embodiment of this application, the disambiguation level of the target region is determined based on the presence of standard target region text and standard jurisdiction region text in the target business processing request. This includes: when the detection result indicates that the business processing request involves administrative division text related to the target region, the similarity calculation is performed between the target standard administrative division text and the target administrative division text included in the target business processing request to obtain a matching similarity; when the matching similarity is greater than or equal to a preset similarity threshold, the disambiguation level is determined to be a first level; when the matching similarity is less than the preset similarity threshold, the disambiguation level is determined to be a second level.

[0084] When the above detection results indicate that the business processing request involves administrative division information related to the target area, it means that the target object is indeed describing the geographical entities related to the target area in this round, but the expression method deviates from the standard division database and fails to meet the aforementioned conditions of precise matching.

[0085] In this case, the next round of fuzzy matching can be initiated, where text similarity is calculated to determine the degree of matching between the target object's description and the target administrative division text. For example, similarity calculation can be performed between the target standard administrative division text and the target administrative division text.

[0086] In practice, a similarity comparison algorithm based on the Longest Common Subsequence (LCS) can be used to calculate the text similarity sim between the two, and this sim can be used as the matching similarity. The formula can be shown in the following formula (1).

[0087] (1)

[0088] Where M is the total number of matching characters between the target standard administrative division text and the target administrative division text, and T is the sum of the total number of characters between the target standard administrative division text and the target administrative division text.

[0089] If the calculated matching similarity is greater than or equal to the preset similarity threshold, it indicates that although the target standard administrative division text and the target administrative division text are not completely identical in form, they are highly similar at the textual level, which is sufficient to confirm that they refer to the same geographical entity. At this time, the disambiguation level can be determined as Level 1.

[0090] If the matching similarity is less than the preset similarity threshold, it indicates that there is a significant difference between the target standard administrative division text and the target administrative division text, and it cannot be confirmed that they point to the same geographic entity. In this case, the disambiguation level can be determined as level two (which can be labeled as guess, corresponding to medium confidence).

[0091] In this embodiment of the application, the aforementioned preset similarity threshold is preferably 0.88. This threshold is determined based on experimental optimization of place name text (typically 4 to 8 characters, including the name of the jurisdiction and the name of the target area), and can balance recall and precision. Meanwhile, each of the above determination steps requires the text fragments participating in the comparison to be at least 2 characters long, in order to reduce the possibility of mismatches caused by single residual characters in the suffix (such as the remaining "state" or "province" after stripping), further improving the reliability of fuzzy matching.

[0092] According to embodiments of this application, the above process can improve the tolerance for non-standard variations (including but not limited to abbreviations, typos, word order differences, suffix omissions, etc.) in business processing requests and improve the mapping accuracy of geographic entities.

[0093] According to an embodiment of this application, the business processing method based on hierarchical disambiguation further includes: in response to the regional query result indicating that the target region and the jurisdictional region do not exist, determining the disambiguation level as the second level.

[0094] When a regional query result indicates that the target region and its jurisdiction do not exist, that is, when the corresponding target region and its jurisdiction cannot be found from the standard regional comparison information based on the target value of the region identifier field in the geographic slot dictionary, it indicates that the geographic entity pointed to by the region identifier field does not exist in the standard regionalization database or has been removed.

[0095] At this point, it is impossible to complete any effective geographic entity mapping based on the target value; therefore, the disambiguation level is determined to be Level 2.

[0096] According to the embodiments of this application, the above processing method can be used as an anomaly protection mechanism to deal with abnormal situations such as asynchronous updates of the standard regionalization library version, data migration, or data errors.

[0097] According to an embodiment of this application, the business processing method based on hierarchical disambiguation further includes: responding to the absence of a target value in the area identifier field of the geographic slot dictionary, querying the target string corresponding to the area name field from the geographic slot dictionary to obtain a name query result; if the name query result indicates the presence of a target string, determining the disambiguation level as the third level; if the name query result indicates the absence of a target string, determining the disambiguation level as the fourth level.

[0098] If the region identifier field in the geoslot dictionary does not contain a target value (i.e., the value of the region identifier field in the geoslot dictionary is empty), you can instead query the region name field (e.g., the city field) in the geoslot dictionary to obtain the corresponding target string and get the name query result. The aforementioned region name field is used to store the region name text (e.g., "City B" or "B").

[0099] If the name query result indicates the presence of the target string (i.e., the region name field is not empty), it means the target object mentions a region name but failed to map it to any region identifier in the standard regional division database. In this case, the disambiguation level can be determined to be level three (which can be identified as gap, corresponding to no match).

[0100] If the name query result indicates that the target string does not exist, i.e., the region name field is empty, it means that the target object has never mentioned the region name. In this case, the disambiguation level can be determined to be level four (which can be marked as none, corresponding to not involved).

[0101] According to the embodiments of this application, by using the existence judgment mechanism based on the region identifier field and region name field in the geographic slot dictionary, different situations can be mapped to the corresponding disambiguation level, providing a clearer level basis for subsequent gating decisions, so that appropriate processing strategies can be adopted under different disambiguation levels.

[0102] Figure 4 A flowchart illustrating the determination of the disambiguation level according to an embodiment of this application is shown.

[0103] like Figure 4 As shown, in the process of determining the disambiguation level, it can be determined whether the area identifier field in the geographic slot dictionary has a target value.

[0104] If a target value exists in the region identifier field, determine whether the standard region comparison information includes a standard administrative division entry corresponding to the target value. This standard administrative division entry must include at least the name of the jurisdiction and the name of the target region. If not, determine the disambiguation level as level two (guess).

[0105] If the predefined existence conditions are not met, determine whether the business processing request includes administrative division text related to the target region. If it does not include it, determine the disambiguation level as Level 1 (exact). If it does include it, perform fuzzy matching.

[0106] For example, it determines whether the matching similarity between the target standard administrative division text and the target administrative division text in the target business processing request is less than a preset similarity threshold. If the matching similarity is greater than or equal to the preset similarity threshold, the disambiguation level is determined to be Level 1 (exact); if the matching similarity is less than the preset similarity threshold, the disambiguation level is determined to be Level 2 (guess).

[0107] In cases where standard administrative division entries are included, determine whether the presence of standard target area text and standard jurisdiction area text in the target business processing request meets preset existence conditions. If the presence meets the preset existence conditions, the disambiguation level is determined to be Level 1 (exact).

[0108] If the region identifier field does not have a target value, further determine whether the region name field in the geographic slot dictionary contains a target string.

[0109] If the target string exists in the region name field, the disambiguation level is determined to be level three (gap). If the target string does not exist in the region name field, it indicates that the user did not involve region information in this round, and the disambiguation level is determined to be level four (none).

[0110] Through the above four-level progressive judgment process, the disambiguation results of geographic entities can be quantified into four progressive disambiguation levels: none, gap, exact, and guess, so as to facilitate subsequent targeted processing.

[0111] According to embodiments of this application, the service processing method based on hierarchical disambiguation further includes: generating a blocking signal and guidance prompt information corresponding to the second, third, or fourth level in response to the disambiguation level being a second level, a third level, or a fourth level; determining natural language guidance in response to the blocking signal; and displaying the natural language guidance in a display interface so that the target object can input a new service processing request or input supplementary information for the service processing request based on the natural language guidance.

[0112] When the disambiguation level of the target area is determined to be Level 2, Level 3, or Level 4, it indicates that the system cannot confirm with sufficient confidence that the target area corresponding to the area identifier field in the current geographic slot dictionary is the geographic entity the user intends to point to. At this time, a blocking signal can be generated, along with guidance information corresponding to the current disambiguation level. This blocking signal is used to prevent subsequent processes from calling the service interface corresponding to the business type, thereby reducing the possibility of passing uncertain or invalid geographic entity parameters to downstream business APIs, and thus reducing business processing anomalies caused by incorrect parameters.

[0113] The strategy for generating the aforementioned guidance and prompts is related to the specific type of disambiguation level. When the disambiguation level is level three, it indicates that the target object mentions a geographic entity but cannot map it to any regional identifier in the standard zoning database. Therefore, corresponding prompts can be generated to prompt the target object to restate its geographic intent using a standard format (such as a standard description that includes the jurisdiction and the target area).

[0114] For example, the guidance message corresponding to the third level could be: The prompt message [City {City Name}] has not yet been found in the standard city database and is not available in any searchable administrative region. Please prompt the user to rephrase the message using the standard [Province + City] format (do not mention the number or internal field name to the user).

[0115] When the disambiguation level is level two (guess), it indicates that a region identifier has been mapped according to the description of the target object, but its confidence level is questionable. Therefore, the corresponding jurisdictional region name and target region name can be obtained from the standard zoning library according to the target value of the region identifier field, and a corresponding confirmation prompt message can be generated. This prompt message can include candidate geographic entities inferred by the system, which are used to request the user to confirm whether they are the geographic entities that the target object intends to point to.

[0116] For example, the guidance message corresponding to the second level could be: Search for the province and city name in the standard library based on the city identifier, and generate a prompt: The standard library entries inferred by the system based on the user's current statement may not be entirely consistent with their original statement. Please ask the user to confirm in natural conversational language, such as: [I guess the city you mentioned is [{province / city}], [Do you mean {province / city}?], and explain that if not, please directly state the correct province and city; do not use internal terms such as [[guessed correctly]] with the user; and prohibit the use of statements such as [You mentioned… Is this city description accurate?] that seem to be restating the user's original statement and questioning its accuracy.

[0117] When the disambiguation level is level four (none), it indicates that the target object's dialogue does not involve any geographic entity information. In this case, corresponding prompts can be generated based on the current business type to guide the user to supplement the missing geographic information.

[0118] All the above guidance and prompts must comply with global constraints, namely, prohibiting the mention of internal field numbers such as area identifiers to users, prohibiting the use of internal technical terms (such as "guess hit", "gap", "guess", etc.), prohibiting the use of expressions that paraphrase the user's original words and question their accuracy, and requiring the use of natural, friendly, and conversational language.

[0119] After generating the blocking signal and guidance prompts, natural language guidance can be determined based on the guidance prompts in response to the blocking signal. The method for determining the natural language guidance is not limited; it can involve inputting the guidance prompts into a large language model, which then generates guidance statements in natural language form based on the prompts. These guidance statements exhibit differentiated expression styles and content for different disambiguation levels. Alternatively, natural language guidance templates corresponding to each guidance prompt can be generated, and the current natural language guidance can be determined through template matching.

[0120] Subsequently, the natural language guidance is displayed to the target object in the interface, allowing the target object to input new business processing requests based on the guidance, or to supplement missing geographic information for the current business processing request. In response to the target object's new business processing requests or supplementary information, semantic parsing and subsequent hierarchical disambiguation steps are re-executed, thereby achieving iterative information completion and geographic entity confirmation in multi-turn dialogues.

[0121] In some embodiments, blocking signals and release signals can be represented by the same Boolean type variable (e.g., block), where a first value (e.g., true) represents a blocking signal and a second value (e.g., false) represents a release signal.

[0122] According to embodiments of this application, when the disambiguation confidence is insufficient, the request is routed to the guiding dialogue node instead of the service interface call node, thereby preventing erroneous parameters from being passed to the downstream business API. This reduces business processing anomalies caused by geographic entity mapping errors, and improves business processing reliability and user experience.

[0123] In some embodiments, the above-described hierarchical disambiguation-based business processing method can be deployed in a dialogue orchestration engine. This dialogue orchestration engine is implemented based on a directed state graph and includes multiple processing nodes and conditional routing edges, including at least a guidance dialogue node and at least one service interface call node. The service interface call node (e.g., an API call node) is used to call downstream service interfaces corresponding to the business type, and the guidance dialogue node is used to output guidance information to the user to confirm or supplement geographical information. In the automatic acquisition node, after the system detects the aforementioned blocking signal, it triggers conditional routing logic, directing the current business processing request to the guidance dialogue node (e.g., a guidance node), rather than any service interface call node. The guidance dialogue node obtains the aforementioned generated guidance prompt information and injects it into the contextual prompts of the large language model, which then outputs guidance statements in natural language form to the user.

[0124] Through the above routing mechanism, when the disambiguation level is not the first level, the business processing request is routed to the guidance dialogue node to guide the user to confirm or supplement information. The request will not reach any service interface call node, thereby isolating the incorrect geographic entity parameters at the system routing level before the downstream service interface call, thus reducing the abnormal calls to downstream service interfaces caused by incorrect geographic entity mapping or insufficient confidence.

[0125] In other words, when the disambiguation result is poor, the request is intercepted at the routing level by the gating mechanism. The conditional routing of the dialogue orchestration engine will direct the request to the guiding dialogue node instead of the service interface call node, and the request will never reach the service interface call node. Through the above routing mechanism, when the disambiguation level is insufficient, business processing requests will be actively routed to the guiding dialogue node, rather than passively waiting for the API call to fail before exception handling, thereby reducing erroneous API calls.

[0126] Figure 5 An application scenario diagram of a service processing method and apparatus based on hierarchical disambiguation according to another embodiment of this application is shown.

[0127] In some embodiments, the hierarchical disambiguation-based business processing method can be implemented jointly by a disambiguation engine, a dialogue orchestration engine, and a large language model. The disambiguation engine may include an administrative division suffix stripping module, a hierarchical disambiguation module, and a hierarchical gating decision module. The dialogue orchestration engine may include slot extraction nodes, flow control nodes, automatic acquisition nodes, guided dialogue nodes, and service interface call nodes.

[0128] like Figure 5 As shown, the slot extraction node in the dialogue orchestration engine can be used to call the large language model for slot extraction, call the geographic hierarchy auto-completion function to automatically complete the target area or jurisdiction level, and merge the extraction results with the current geographic slot dictionary.

[0129] The large language model is responsible for slot extraction, which involves extracting structured slot fields (such as region identifier and region name fields) from the business processing requests of the target object and outputting the geographic slot dictionary to the disambiguation engine.

[0130] The flow control node in the dialogue orchestration engine is used to invoke the mandatory slot check logic to determine whether the mandatory fields required for the current business model are complete; to invoke the hierarchical gating decision module to perform city entity disambiguation and generate guidance prompts; and to invoke the large language model to obtain intent and determine whether the user wants to obtain the analysis results immediately (or can combine LLM and regular expression heuristics to determine this). The automatic acquisition node is used to invoke the hierarchical gating decision module to obtain the final disambiguation result and set routing signals based on blocking or allowing signals.

[0131] For example, when the hierarchical gating decision module outputs block=True, the automatic acquisition node detects the blocking marker, and the conditional routing directs the business processing request to the guidance dialogue node. The guidance dialogue node obtains the aforementioned guidance prompt information and injects it into the contextual prompts of the large language model. The large language model then outputs guidance statements in natural language form to the user to guide the user to confirm or supplement geographical information.

[0132] When the hierarchical gating decision module outputs block=False, the conditional routing directs the business processing request to the service interface call node. This node is a middleware that encapsulates the downstream business service interface call, used to route to the corresponding service interface based on the business type, obtain business data, summarize the data, and end the current processing flow.

[0133] In some embodiments, the automatically acquired node can also be used to perform geographic hierarchy autocomplete and mandatory field checks again, and to call the hierarchical gating decision module again to obtain the final disambiguation result, and set the routing signal according to the blocking signal or the allowing signal.

[0134] The administrative division suffix stripping module in the disambiguation engine performs the suffix stripping operation. Specifically, this module strips the administrative division text included in the business processing request and the corresponding standard administrative division text in the standard administrative division library, respectively, according to the longest priority principle, to obtain the target business processing request and the target standard administrative division text. The city-level disambiguation module performs a four-level confidence judgment based on the stripped text, quantifying the entity disambiguation result into four progressive disambiguation levels: none, gap, exact, and guess. The input of this module includes at least a geographic slot dictionary and a standard administrative division library lookup table (or standard area lookup information), and the output includes at least the disambiguation level and the corresponding guidance prompts.

[0135] The hierarchical gating decision module is used to perform routing decisions based on disambiguation level and service type. Specifically, it can output blocking or allowing signals (e.g., represented by the boolean variable block, where block=True indicates blocking and block=False indicates allowing) and a corresponding guidance prompt list according to the rule of "block, guess, or / none" (block, allow, allow).

[0136] Through the above module division and routing mechanism, the disambiguation engine runs independently of the large language model and downstream service interfaces. Through the above judgment logic, the same input can be executed multiple times to obtain the same disambiguation result, which solves the randomness and uncontrollability problems of related technologies.

[0137] Furthermore, through modular decoupling among the three components, each module can be tested, optimized, and replaced independently. They can also interact and call each other through standardized interfaces.

[0138] In addition, the state dictionary of the dialogue orchestration engine can contain the following key fields related to disambiguation gating: The first field (Boolean value), set by the automatic acquisition node based on the return result of the hierarchical gating decision module, is used to mark whether the current business processing request needs to be blocked due to city confirmation or the standard zoning database not being filed. This field can determine whether to enter the service interface call node. The second field (string), a pre-built guidance prompt message, contains complete context information such as the current business type, the collected geographic slot dictionary snapshot, and disambiguation prompt lines, which is consumed by the large language model and output as user guidance. The third field (string list), a list of disambiguation prompt lines, where each line corresponds to a specific disambiguation question (such as city gap or city guess), for the guidance dialogue node to inject prompts into the large language model.

[0139] Furthermore, the construction and maintenance of the standard zoning database can be performed as follows: Online administrative division records (i.e., those that can be processed) can be queried from the administrative division climate information table stored in the database. The query results are then formatted into a three-column text format: "Regional Identifier, Jurisdiction Area Name, and Target Area Name." Multiple such text formats are then stored in the database as the standard zoning database. At the start of each dialogue round, the latest standard zoning database can be retrieved from the database, and a cache validity period for the target duration can be set. If the database is unavailable, static zoning data from a local snapshot file can be used as a fallback to ensure service availability.

[0140] Based on the above-described hierarchical disambiguation-based service processing method, this application also provides a hierarchical disambiguation-based service processing apparatus. The following will combine... Figure 6 The device is described in detail.

[0141] Figure 6 A structural block diagram of a service processing apparatus based on hierarchical disambiguation according to an embodiment of this application is shown.

[0142] like Figure 6 As shown, the hierarchical disambiguation-based service processing device 600 includes a parsing module 610, a stripping module 620, a query module 630, a determination module 640, and a processing module 650.

[0143] The parsing module 610 is used to perform semantic parsing on the business processing request in response to receiving the business processing request, and to determine the geographic slot dictionary and business type corresponding to the business processing request; in one embodiment, the parsing module 610 can be used to perform the operation S210 described above, which will not be repeated here.

[0144] The stripping module 620 is used to strip the administrative division text included in the business processing request and the standard administrative division text corresponding to the administrative division text in the standard administrative division library according to the longest priority principle when the business type is the target business type, so as to obtain the target business processing request and the target standard administrative division text. In one embodiment, the stripping module 620 can be used to perform the operation S220 described above, which will not be repeated here.

[0145] The query module 630 is used to query the target region corresponding to the target value and the jurisdiction to which the target region belongs from the standard region comparison information in response to the existence of a target value in the region identifier field of the geographic slot dictionary, and obtain the region query result; in one embodiment, the query module 630 can be used to perform the operation S230 described above, which will not be repeated here.

[0146] The determining module 640 is used to determine the disambiguation level of the target region based on the matching information between the target business processing request and the target standard administrative division text in response to the regional query result indicating the existence of the target region and the jurisdiction region. In one embodiment, the determining module 640 can be used to perform the operation S240 described above, which will not be repeated here.

[0147] The processing module 650 is configured to, in response to the disambiguation level being the first level, invoke the service interface corresponding to the service type to process the service processing request, so that the service interface generates a service processing result corresponding to the target region. In one embodiment, the processing module 650 may be used to execute the operation S250 described above, which will not be repeated here.

[0148] According to embodiments of this application, any multiple modules among the parsing module 610, stripping module 620, query module 630, determination module 640, and processing module 650 can be combined into one module, or any one of these modules can be split into multiple modules. Alternatively, at least part of the functionality of one or more of these modules can be combined with at least part of the functionality of other modules and implemented in one module. According to embodiments of this application, at least one of the parsing module 610, stripping module 620, query module 630, determination module 640, and processing module 650 can be at least partially implemented as hardware circuitry, such as a field-programmable gate array (FPGA), a programmable logic array (PLA), a system-on-a-chip, a system-on-a-substrate, a system-on-package, an application-specific integrated circuit (ASIC), or implemented in hardware or firmware by any other reasonable means of integrating or packaging circuitry, or implemented in software, hardware, or firmware, or in any suitable combination of any of these three implementation methods. Alternatively, at least one of the parsing module 610, stripping module 620, query module 630, determination module 640 and processing module 650 may be implemented at least partially as a computer program module, which can perform corresponding functions when the computer program module is run.

[0149] Figure 7 A block diagram of an electronic device suitable for implementing a hierarchical disambiguation-based service processing method according to an embodiment of this application is shown.

[0150] like Figure 7 As shown, an electronic device 700 according to an embodiment of this application includes a processor 701, which can perform various appropriate actions and processes according to a program stored in a read-only memory ROM 702 or a program loaded from a storage portion 708 into a random access memory RAM 703. The processor 701 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and / or an associated chipset and / or a special-purpose microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 701 may also include onboard memory for caching purposes. The processor 701 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of this application.

[0151] RAM 703 stores various programs and data required for the operation of electronic device 700. Processor 701, ROM 702, and RAM 703 are interconnected via bus 704. Processor 701 executes various operations of the method flow according to embodiments of this application by executing programs in ROM 702 and / or RAM 703. It should be noted that the programs may also be stored in one or more memories other than ROM 702 and RAM 703. Processor 701 may also execute various operations of the method flow according to embodiments of this application by executing programs stored in said one or more memories.

[0152] According to embodiments of this application, the electronic device 700 may further include an input / output (I / O) interface 705, which is also connected to a bus 704. The electronic device 700 may also include one or more of the following components connected to the input / output (I / O) interface 705: an input section 706 including a keyboard, mouse, etc.; an output section 707 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 708 including a hard disk, etc.; and a communication section 709 including a network interface card such as a LAN card, modem, etc. The communication section 709 performs communication processing via a network such as the Internet. A drive 710 is also connected to the input / output (I / O) interface 705 as needed. A removable medium 711, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on the drive 710 as needed so that computer programs read from it can be installed into the storage section 708 as needed.

[0153] This application also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiments; or it may exist independently and not assembled into the device / apparatus / system. The computer-readable storage medium carries one or more programs, which, when executed, implement the method according to the embodiments of this application.

[0154] According to embodiments of this application, the computer-readable storage medium can be a non-volatile computer-readable storage medium, such as including but not limited to: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this application, the computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. For example, according to embodiments of this application, the computer-readable storage medium may include ROM 702 and / or RAM 703 and / or one or more memories other than ROM 702 and RAM 703 described above.

[0155] Embodiments of this application also include a computer program product comprising a computer program containing program code for performing the methods shown in the flowchart. When the computer program product is run on a computer system, the program code is used to cause the computer system to implement the methods provided in the embodiments of this application.

[0156] When the computer program is executed by the processor 701, it performs the functions defined in the system / apparatus of this application embodiment. According to the embodiments of this application, the systems, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0157] In one embodiment, the computer program may rely on a tangible storage medium such as an optical storage device or a magnetic storage device. In another embodiment, the computer program may also be transmitted and distributed in the form of signals over a network medium, and may be downloaded and installed via the communication section 709, and / or installed from a removable medium 711. The program code contained in the computer program can be transmitted using any suitable network medium, including but not limited to: wireless, wired, etc., or any suitable combination thereof.

[0158] In such an embodiment, the computer program can be downloaded and installed from a network via the communication section 709, and / or installed from the removable medium 711. When the computer program is executed by the processor 701, it performs the functions defined in the system of this application embodiment. According to the embodiments of this application, the systems, devices, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0159] According to embodiments of this application, program code for executing the computer programs provided in the embodiments of this application can be written in any combination of one or more programming languages. Specifically, these computational programs can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. Programming languages ​​include, but are not limited to, languages ​​such as Java, C++, Python, "C", or similar programming languages. The program code can be executed entirely on the user's computing device, partially on the user's device, partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0160] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0161] Those skilled in the art will understand that the features described in the various embodiments of this application can be combined and / or combined in various ways, even if such combinations or combinations are not explicitly described in this application. In particular, the features described in the various embodiments of this application can be combined and / or combined in various ways without departing from the spirit and teachings of this application. All such combinations and / or combinations fall within the scope of this application.

[0162] The embodiments of this application have been described above. However, these embodiments are merely illustrative and not intended to limit the scope of this application. Although various embodiments have been described above, this does not mean that the measures in the various embodiments cannot be used advantageously in combination. Without departing from the scope of this application, those skilled in the art can make various substitutions and modifications, all of which should fall within the scope of this application.

Claims

1. A business processing method based on hierarchical disambiguation, characterized in that, The method includes: In response to receiving a business processing request, the system performs semantic parsing on the business processing request to determine the geoslot dictionary and business type corresponding to the business processing request. When the business type is the target business type, according to the longest priority principle, the administrative division text included in the business processing request and the standard division text in the standard division library corresponding to the administrative division text are stripped of their division suffixes to obtain the target business processing request and the target standard administrative division text. In response to the existence of a target value in the region identifier field of the geographic slot dictionary, the target region corresponding to the target value and the jurisdiction to which the target region belongs are queried from the standard region comparison information to obtain the region query result; In response to the regional query results indicating the existence of the target region and the jurisdiction, the disambiguation level of the target region is determined based on the matching information between the target business processing request and the target standard administrative division text. In response to the disambiguation level being the first level, the service interface corresponding to the business type is invoked to process the business processing request, so that the service interface generates a business processing result corresponding to the target region.

2. The business processing method according to claim 1, characterized in that, In response to the disambiguation level being the first level, the service interface corresponding to the business type is invoked to process the business processing request, so that the service interface generates a business processing result corresponding to the target region, including: In response to the disambiguation level being the first level, a release signal corresponding to the business processing request and a call request for invoking the service interface corresponding to the business type are generated, wherein the call request includes text corresponding to the target region and the jurisdiction region concatenated according to a preset format; In response to the release signal, the corresponding service interface is invoked based on the call request, so that the service interface generates a business processing result corresponding to the target area.

3. The business processing method according to claim 1 or 2, characterized in that, The step of semantically parsing the business processing request to determine the geoslot dictionary and business type corresponding to the business processing request includes: The business processing request is semantically parsed to obtain the parsing result; If the parsing result indicates that the business processing request does not include administrative division text, a historical geographic slot dictionary corresponding to the target object is determined based on the target object that sent the business processing request; Use the historical geographic slot dictionary as the geographic slot dictionary corresponding to the business processing request; Based on the parsing results, the business type corresponding to the business processing request is determined.

4. The business processing method according to claim 1, characterized in that, The target standard administrative division text includes the standard target area text corresponding to the target area and the standard jurisdiction area text corresponding to the jurisdiction area; The step of determining the disambiguation level of the target region based on the matching information between the target business processing request and the target standard administrative division text includes: The target business processing request is compared with the target standard administrative division text to determine the presence of the standard target area text and the standard jurisdiction area text in the target business processing request; Based on the presence of the standard target region text and the standard jurisdiction region text in the target service processing request, the disambiguation level of the target region is determined.

5. The business processing method according to claim 4, characterized in that, The step of determining the disambiguation level of the target region based on the presence of the standard target region text and the standard jurisdiction region text in the target service processing request further includes: If the existence condition satisfies the preset existence conditions, the disambiguation level is determined to be the first level; The preset existence conditions include at least one of the following: The standard target region text exists in the target business processing request; and, The standard target region text and the standard jurisdiction region text both exist in the target business processing request.

6. The business processing method according to claim 4, characterized in that, The step of determining the disambiguation level of the target region based on the presence of the standard target region text and the standard jurisdiction region text in the target service processing request includes: If the existence condition does not meet the preset existence conditions, based on the existence of the standard target area text and the standard jurisdiction area text in the target business processing request, and the existence of the division suffix in the business processing request, the business processing request is subjected to area involvement detection to obtain the detection result; If the detection result indicates that the business processing request does not involve administrative division text related to the target area, the disambiguation level is determined to be the first level.

7. The business processing method according to claim 6, characterized in that, The step of determining the disambiguation level of the target region based on the presence of the standard target region text and the standard jurisdiction region text in the target service processing request includes: If the detection result indicates that the business processing request involves administrative division text related to the target region, the similarity between the target standard administrative division text and the target administrative division text included in the target business processing request is calculated to obtain the matching similarity. If the matching similarity is greater than or equal to a preset similarity threshold, the disambiguation level is determined to be the first level; If the matching similarity is less than a preset similarity threshold, the disambiguation level is determined to be the second level.

8. The business processing method according to claim 1, characterized in that, The method further includes: In response to the regional query result indicating that the target region and the jurisdictional region do not exist, the disambiguation level is determined to be the second level.

9. The business processing method according to claim 1, characterized in that, The method further includes: In response to the absence of a target value in the region identifier field of the geographic slot dictionary, the target string corresponding to the region name field is queried from the geographic slot dictionary to obtain the name query result; If the name query results indicate that the target string exists, the disambiguation level is determined to be level three; If the name query result indicates that the target string does not exist, the disambiguation level is determined to be the fourth level.

10. The business processing method according to any one of claims 1, 7, 8, and 9, characterized in that, The method further includes: In response to the disambiguation level being the second, third, or fourth level, a blocking signal and guidance prompt information corresponding to the second, third, or fourth level are generated; In response to the blocking signal, natural language guidance is determined based on the guidance prompt information; The natural language guidance is displayed in the display interface so that the target object can input new business processing requests or input supplementary information for the business processing requests based on the natural language guidance.

Images