A government affair matter self-adaptive verification method based on subject fact state

Abstract

This invention relates to the field of structured data processing technology for government affairs, and discloses an adaptive verification method for government affairs based on the factual state of the subject. The method converts structured data from at least two government data sources into subject factual state units and stores them in a state database. Upon receiving a verification request, it generates factual requirement data decoupled from the specific data source interface. Before generating the data source access task, it matches candidate factual state units and dynamically determines the applicable verification state based on the objective factual state, factual time information, and the current proof requirement. When directly applicable, candidate facts are reused; when not directly applicable, data processing actions are triggered and the state database is updated. Simultaneously, the verification execution trajectory is recorded, and when discrepancies are verified, discrepancy nodes are located. Only the verification strategy parameters are adjusted without changing the factual judgment criteria and logical relationships.

Description

Technical Field

[0001] This invention relates to the fields of government structured data processing, heterogeneous data processing, and data status management, specifically to an adaptive verification method, system, electronic device, and computer-readable storage medium for government matters based on the factual status of the subject. Background Technology

[0002] With the development of digital government and online government services, an increasing number of government affairs require automatic verification based on multiple government data sources. When handling matters such as qualification confirmation, document verification, industry access, subsidy application, housing security, and health certificates, government service systems typically need to access multiple data sources, including population, market entities, real estate, housing, social insurance, permits and approvals, and licenses, to determine whether the applicant or business entity meets the requirements.

[0003] Existing automated government verification systems typically configure data interfaces, field mappings, and review rules around specific matters. Upon receiving a verification request, the system accesses the corresponding data source according to a preset process, obtains structured data, and then makes a judgment based on the matter's rules. While this approach can support automated verification of a single matter, it easily leads to problems such as repeated data source calls, repeated field conversions, and repeated structured data processing when verifying the same fact for multiple matters, increasing interface access pressure and system processing costs.

[0004] Meanwhile, different government data sources originated from different business systems or historical stages, and their data standards, update times, status meanings, and object identifiers are not entirely consistent. The existence of raw data does not necessarily mean that the data can serve as valid proof for the current matter. For example, a historical fact may have been true at a specific time, but may have become invalid within the time frame required by the current target matter; business objects in one data source and business objects in another data source may describe the same real-world object, or may not be directly confirmed due to historical migration, missing fields, or inconsistent identifiers. If the system makes verification conclusions based solely on interface return values ​​or the existence status of a single data source, it may lead to the misuse of historical facts, failure to identify conflicting facts, or incorrect acceptance of facts to be confirmed.

[0005] Furthermore, some facts are not directly derived from a single data source, but rather need to be calculated and generated based on source facts, directory mappings, classification rules, or derivation strategies. When the source facts, directory versions, mapping relationships, or derivation strategies change, existing derived facts may become invalid. If the system performs a full recalculation of all subject data, it will result in a large amount of irrelevant calculations; if it does not recalculate, invalid derived facts may continue to be used for verification.

[0006] When discrepancies exist between automated verification results and manual review results or other reliable review results, existing systems typically only record the review conclusion or directly modify the final judgment rule. This approach makes it difficult to pinpoint whether the discrepancy stems from data time, data conflicts, missing evidence links, or invalid derived facts. It also makes it easy for verification strategy adjustments to deviate from the original factual judgment requirements and logical relationships, affecting the consistency of verification standards for similar matters.

[0007] Even if some systems have established shared data pools, they usually synchronize data according to fixed periods or call data according to the process of the task interface, and cannot dynamically judge the applicability of existing facts based on the current verification needs; the review feedback is also mostly used to modify the final review rules or manual conclusions, and it is difficult to settle into controlled verification strategy parameters without changing the factual determination elements and their logical relationships.

[0008] Therefore, there is a need for an adaptive verification method that can convert multi-source government data into reusable subject fact state units, prioritize whether existing facts meet the current fact verification requirements before generating data source access tasks, and adjust only the fact application and data processing related parameters without changing the boundary of fact determination rules under the feedback of verification differences. Summary of the Invention

[0009] The purpose of this invention is to provide an adaptive verification method for government affairs based on the factual status of the subject, so as to solve the problems of repeated access to multi-source data, misuse of historical facts, difficulty in identifying cross-source fact conflicts, high cost of full recalculation after changes in derived facts, and uncontrollable accumulation of review feedback in existing automatic government affairs verification.

[0010] To achieve the above objectives, this invention provides an adaptive verification method for government affairs based on subject factual status. This method acquires structured data from at least two government data sources, converts the structured data into standardized subject factual status units, and stores them in a subject factual status database. Each subject factual status unit represents the factual status of the subject to be verified, related subjects, or business objects within a specific time frame, and includes at least subject identifier, fact type, fact value, fact time information, data source identifier, and objective factual status.

[0011] Upon receiving a verification request for a target government matter, the system does not directly perform verification according to the specific government data source interface. Instead, it first generates a fact-verification requirement based on the verification request. This fact-verification requirement represents the facts that need to be proven for the current target government matter. It includes at least the subject to be verified, the type of fact to be proven, the time constraint, and the status of the target fact, but does not include the specific government data source interface identifier or original field mapping information. This approach decouples the fact-verification target from the underlying specific data acquisition interface.

[0012] Before generating a data source access task, the system prioritizes matching candidate subject fact state units corresponding to the factual verification requirements from the subject fact state database. For each matched candidate subject fact state unit, the system dynamically determines the verification applicability status of the candidate subject fact state unit in the current target government matter based on its objective factual status, factual time information, and current factual verification requirements.

[0013] When the verification applicability status is "directly applicable," the system directly reuses the corresponding subject fact state unit to generate the factual result, and does not generate a data source access task for the factual requirement. When the verification applicability status is not "directly applicable," the system triggers the corresponding data processing action according to the verification applicability status, and updates the subject fact state database according to the processing result.

[0014] The system generates automatic verification results based on the subject factual state units that meet the factual verification requirements, and records the verification execution trajectory. The verification execution trajectory records the relationships between factual verification requirements, candidate subject factual state units, applicable verification status, data processing actions, data processing results, and automatic verification results. When a discrepancy exists between the review result and the automatic verification result, the system traces the factual verification node causing the discrepancy back along the verification execution trajectory. It only adjusts the verification strategy parameters affecting the applicable verification status or data processing actions, without changing the factual determination requirements and their logical relationships corresponding to the target government matter.

[0015] Compared with the prior art, the present invention has at least the following beneficial effects: First, the present invention converts multi-source government data into reusable data state expressions through subject fact state units and subject fact state libraries. When multiple target government matters need to verify the same or similar facts, existing subject fact state units can be reused first, reducing repeated data source calls and repeated structured processing. Second, this invention decouples the factual requirement from the specific data source interface. The system no longer relies solely on fixed matter interface processes for verification, but instead performs candidate fact matching and applicability status judgment based on the facts that need to be proved for the current target government matter, thereby improving the flexibility and maintainability of multi-source government data verification. Third, the present invention dynamically determines the applicable verification status based on the objective factual state, factual time information, and current factual requirements, and can distinguish between currently valid facts, historically valid facts, conflicting facts, expired derived facts, and facts with insufficient proof, thereby reducing the risk of historical facts and expired data being incorrectly reused. Fourth, this invention determines the set of affected derived facts through version dependency information and reverse indexes. When the source facts, directory version, mapping relationship or derivation strategy changes, it only performs recalculation or sets a recalculation mark for the set of affected derived facts, thus avoiding full processing of irrelevant subject facts. Fifth, when there is a difference between the review result and the automatic verification result, the present invention locates the difference node along the verification execution trajectory and restricts the strategy adjustment to the range of parameters that affect the applicable state of verification or data processing actions. By constraining invariance verification, it ensures that the fact judgment elements and their logical relationships are not changed, thereby improving the consistency and traceability of the review feedback sedimentation process. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the overall process of an adaptive verification method for government affairs based on the factual status of the subject, provided in an embodiment of the present invention.

[0017] Figure 2 A schematic diagram of the dynamic judgment process for the reuse and verification of the subject fact state provided in the embodiments of the present invention.

[0018] Figure 3 This is a schematic diagram of the reverse location process for verification execution trajectory and review difference provided in an embodiment of the present invention.

[0019] Figure 4 This is a schematic diagram of the constraint invariance verification process provided in an embodiment of the present invention. Detailed Implementation

[0020] The technical solution of the present invention will be described below with reference to embodiments. It should be understood that the described embodiments are only for explaining the present invention and are not intended to limit the scope of protection of the present invention. Without departing from the principle of the present invention, those skilled in the art can adjust the data source type, fact type, state classification, processing action, storage method and judgment order.

[0021] This embodiment provides an adaptive verification method for government affairs based on the factual status of the subject. It can be applied to government affairs processing systems, unified acceptance platforms, business approval systems, data sharing and exchange platforms, or other information systems that require automatic verification based on multi-source government data. The system can obtain structured data from different government data sources through government data sharing interfaces, database synchronization interfaces, file exchange interfaces, or message queue interfaces.

[0022] In this embodiment, government data sources may include population data sources, market entity data sources, real estate data sources, commercial housing contract filing data sources, certificate data sources, permit approval data sources, social insurance data sources, item rule data sources, directory data sources, and other data sources that can provide structured data. The system performs field standardization, subject identification processing, time information extraction, fact type mapping, and status mapping on the data returned by each government data source, converting it into subject fact state units.

[0023] The subject fact state unit is used to describe the fact state of a subject, related subject, or business object within a specific time frame. In one implementation, the subject fact state unit can be represented by the following data structure: FactUnit = {subject_id, fact_type, fact_value, related_subject,source_id, event_time, update_time, objective_status, version_info} In this structure, `subject_id` represents the subject identifier; `fact_type` represents the fact type; `fact_value` represents the fact value; `related_subject` represents the associated subject or object; `source_id` represents the data source identifier; `event_time` represents the time the fact occurred; `update_time` represents the time the fact was updated; `objective_status` represents the objective fact status; and `version_info` represents version dependency information. The above data structure is merely an example; those skilled in the art can implement it using relational tables, key-value structures, graph structures, or object structures depending on the actual storage method.

[0024] Objective factual status refers to the objective state of a subject's factual state unit in its original data source or during its derivation process. Objective factual status can include currently valid, historically valid, pending confirmation, conflicting, invalid, and unknown. Currently valid indicates that the corresponding fact can be proven to be true under the conditions of the data source and time; historically valid indicates that the corresponding fact was true at a historical point in time, but whether it applies to the current objective matter requires further judgment; pending confirmation indicates that it needs to be confirmed through other data sources or verification methods; conflicting indicates that the fact is inconsistent with related facts in other data sources; invalid indicates that the fact is clearly no longer applicable; unknown indicates that its status cannot be determined at present.

[0025] The main fact state unit can include source fact state units and derived fact state units. Source fact state units are directly generated from structured data in government data sources, such as the household registration status of natural persons, the registration status of market entities, the registration status of houses, and the validity status of certificates. Derived fact state units are calculated and generated from one or more source fact state units or other derived fact state units, such as categorized facts generated based on source facts and directory mapping relationships, current state facts generated based on multiple time-series facts, and proof link facts generated based on multiple subject relationships.

[0026] For derived fact state units, the system records version dependency information. Version dependency information may include at least one of the following: source fact version, directory version, mapping version, target category directory version, and derivation strategy version. The system establishes a reverse index between the derived fact state unit and its dependent source facts, directory nodes, mapping relationships, or derivation strategies. When a dependency version changes, the system determines the affected set of derived facts based on the reverse index and performs recalculation only on the affected set of derived facts, or sets a recalculation flag for that set.

[0027] The system saves the transformed principal fact state units to the principal fact state repository. The principal fact state repository can be implemented using a relational database, graph database, document database, cache database, or a combination thereof. The principal fact state repository not only stores individual fact states, but also the relationships between principals, between principals and business objects, and between facts, to support subsequent fact-checking for requirement matching and cross-source conflict verification.

[0028] When the system receives a verification request for a target government matter, it first extracts verification semantic elements from the request, such as the subject to be verified, the type of fact to be proven, the state of the target fact, time constraints, and factual integrity requirements, and then converts them into factual requirement. Factual requirement describes the facts that the current target government matter needs to prove, rather than describing which specific data source to access or which specific field to read.

[0029] In one implementation, it has been found that requirements can be represented using the following data structure: ProofNeed = {target_subject, required_fact_type, target_fact_state,time_constraint, relation_path, completeness_requirement} Here, `target_subject` represents the subject to be verified; `required_fact_type` represents the type of fact to be proven; `target_fact_state` represents the state of the target fact; `time_constraint` represents the time constraint; `relation_path` represents the subject relationship path; and `completeness_requirement` represents the proof completeness requirement. It turns out that the requirements do not include specific government data source interface identifiers, nor do they include original field mapping information.

[0030] After generating a fact-verifying requirement, the system prioritizes matching candidate subject fact state units corresponding to the requirement from the subject fact state database before generating a data source access task. The matching process can be performed based on subject identifier, fact type, time range, associated subject, business object identifier, location attribute, or other standardized attributes. If no candidate subject fact state unit exists in the subject fact state database, the system can determine that the applicable verification status needs to be supplemented and trigger the corresponding data source access task.

[0031] When a candidate subject fact state unit exists, the system does not directly reuse the candidate fact. Instead, it dynamically determines the verification applicability state based on the current factual requirements. The verification applicability state indicates the applicability of the candidate subject fact state unit in the current target administrative matter. This verification applicability state is not a globally fixed state of the subject fact state unit. The same subject fact state unit can be determined to have different verification applicability states under different target administrative matters, different time constraints, or different factual requirements.

[0032] In one implementation, the system determines the verification applicability status in the following order: First, determine whether the fact type of the candidate subject's fact state unit meets the current proof requirements. If the fact type does not match, it is determined to be insufficient or inapplicable proof. Second, determine whether the factual time information of the candidate subject's factual state unit meets the time constraints corresponding to the current factual proof requirements. If the factual time has expired, the update time is earlier than the specified deadline, or it cannot cover the target time range, it is determined that it needs to be refreshed or that the proof is insufficient. Third, determine whether the factual state unit of the candidate subject conflicts with the factual state of the same subject or the same business object in other government data sources. If a conflict exists, it is determined that a conflict exists, and cross-source conflict verification is triggered. Fourth, when the candidate subject fact state unit is a derived fact state unit, determine whether its version dependency information is invalid. If the source fact version, directory version, mapping version, or derivation strategy version has changed, it is determined that it needs to be recalculated. Fifth, determine whether the candidate subject's factual state unit can form a complete proof link that satisfies the current factual requirements. If necessary factual nodes are missing, subject relationship paths are incomplete, or factual completeness is insufficient, it is determined that the proof is insufficient or needs to be supplemented. Sixth, when there are no cases of fact type mismatch, time non-compliance, cross-source conflict, dependency failure, or incomplete proof link, and the objective fact state of the candidate subject fact state unit meets the current fact-proving requirements, its verification application state is determined to be directly applicable.

[0033] When multiple abnormal situations coexist, the system can determine the final verification application status according to preset status priorities. In one implementation, the preset status priorities are: conflict is higher than needing recalculation, needing recalculation is higher than needing refresh, needing refresh is higher than insufficient proof, and insufficient proof is higher than direct application. By setting status priorities, inconsistent judgments can be avoided when multiple abnormal conditions coexist.

[0034] When the verification application status is "directly applicable," the system directly reuses the corresponding candidate subject fact state unit from the subject fact state database to generate the factual verification result, without generating a data source access task for this factual verification requirement. This approach can reduce repeated access to the underlying government data source, lower the number of interface calls, and reduce the cost of repetitive structured processing.

[0035] When the verification applicability status requires a refresh, the system determines the corresponding data acquisition capability based on the type of fact to be proven, and accesses the relevant government data source to obtain the latest structured data. The system then converts the refreshed data back into a subject fact state unit and writes it into the subject fact state database. After the refresh is complete, the system re-determines the verification applicability status of that subject fact state unit under the current factual requirement.

[0036] When the verification status indicates a conflict, the system performs cross-source conflict verification. Cross-source conflict verification can be performed based on standardized attributes such as subject identifier, unified social credit code, document number, business object number, location attribute, associated subject attribute, and time attribute. The system determines the status of confirmed association, candidate association, conflicting association, or unresolved association based on the object association and time sequence relationships between facts from multiple sources. For facts confirmed to be in conflict, the system can generate a review task or trigger supplementary data retrieval.

[0037] When the verification status indicates that recalculation is required, the system determines the set of affected derived facts based on the version dependency information and inverted index of the derived fact state unit. The system only performs recalculation on the affected set of derived facts, or triggers delayed recalculation based on the recalculation flag when the derived fact state unit is proven to match the requirements again. In this way, the system can avoid performing a full reprocessing of all subject data when the directory, mapping relationship, or derivation strategy changes.

[0038] When the verification status indicates insufficient evidence or the need for supplementary data acquisition, the system determines the missing fact nodes based on the proven needs and generates a supplementary data acquisition task or a review task. The supplementary data acquisition task can access other government data sources or receive structured results of supplementary materials. The review task provides the reviewers with the proven needs, candidate subject factual status units, missing nodes, conflict information, and the system's judgment results.

[0039] The system generates automatic verification results based on the subject factual state units that meet the factual verification requirements. Automatic verification results can include states such as pass, fail, pending review, and data supplementation required. The system also records the verification execution trajectory. The verification execution trajectory includes the factual verification requirements, candidate subject factual state units, applicable verification states, data processing actions, data processing results, and the relationships between the automatic verification results. The verification execution trajectory can be stored using a node and edge structure, or as a structured log or relational table.

[0040] In one implementation, the verification execution trajectory includes a proof requirement node, a candidate fact node, an applicable status node, a processing action node, a processing result node, and a verification result node. The proof requirement node records the facts required to prove the target government matter; the candidate fact node records the factual status units of the subjects participating in the verification; the applicable status node records the dynamically determined verification applicable status; the processing action node records actions such as fact reuse, data refresh, cross-source conflict verification, recalculation, or supplementary acquisition; the processing result node records the processing action output; and the verification result node records the final automatic verification result.

[0041] When discrepancies exist between the review results and the automatic verification results, the system reverse-engineers the verification execution path. Starting from the automatic verification result node, the system traverses backwards to the processing result node, processing action node, applicable state node, candidate fact node, and proof requirement node to identify the fact verification node that caused the discrepancy between the automatic verification result and the review result. Sources of discrepancies may include fact time invalidation, cross-source fact conflicts, derived fact dependency invalidation, missing proof links, or incompatible state judgment parameters.

[0042] Based on the difference node and the difference source, the system only adjusts the verification strategy parameters that affect the applicable verification status or data processing actions. For example, when the difference source is an expired fact time, the refresh conditions for the corresponding fact type can be adjusted; when the difference source is an expired derived fact dependency, the trigger conditions for the recalculated tag can be adjusted; when the difference source is an unidentified cross-source conflict, the trigger conditions for cross-source conflict verification can be adjusted. The system does not directly change the factual determination requirements and their logical relationships corresponding to the target government affairs matter.

[0043] Before adjusting the verification strategy parameters, the system performs a constraint invariance check. This check confirms that the strategy adjustment does not alter the factual elements and logical relationships of the target government matter. In one implementation, the system converts the factual elements and logical relationships before adjustment into a first normalized constraint representation, and converts the adjusted factual elements and logical relationships into a second normalized constraint representation. When the first normalized constraint representation is consistent with the second normalized constraint representation, and the adjusted parameter belongs to the set of verification strategy parameters that are allowed to be adjusted, the constraint invariance check is deemed successful, and the adjusted verification strategy parameters are allowed to be used for subsequent verifications.

[0044] In one implementation, the normalized constraint representation includes fact-determining node and logical relation node. The fact-determining node can be converted into a unified format representation according to a preset data format; for logical relation nodes that satisfy exchange relationships, their corresponding child nodes can be sorted according to node identifier, fact type, target fact state, or other deterministic sorting rules. Through this normalization process, misjudgments caused by different expression orders can be reduced, improving the stability of constraint invariance verification.

[0045] The set of verification strategy parameters that can be adjusted as described above may include fact refresh time parameters, derived fact recalculation trigger parameters, cross-source conflict verification trigger parameters, and proof link integrity check parameters. This set does not include parameters used to change the fact-determination requirements of the target government matter, the fact-determination operation relationship, or the logical relationship between fact-determination requirements. In this way, the system can maintain the stability of the matter verification rule boundaries while absorbing review feedback.

[0046] In one specific embodiment, the target government matter requires verification of whether a valid relationship exists between the applicant and a specific property within the time frame required by the matter. The system obtains structured data from data sources including commercial housing contract filing, real estate registration, and transaction transfer, converts it into contract filing factual status units, real estate registration factual status units, and transaction transfer factual status units, respectively, and writes them into the subject factual status database.

[0047] The contract filing fact status unit can record the contract number, buyer identifier, property location, filing time, filing status, and data source identifier; the real estate registration fact status unit can record the registration unit number, rights holder identifier, property location, registration time, rights status, and data source identifier; the transaction transfer fact status unit can record the transferring parties, transfer time, transfer type, and corresponding property information. Upon receiving a verification request, the system generates a fact-verification requirement, indicating the need to prove a valid relationship between the applicant or related parties and a specific property within a target timeframe.

[0048] The system first matches candidate subject fact state units in the subject fact state database. When only historical contract filing facts are matched, the system determines whether the time information of this fact meets the current target time constraint and whether there are subsequent real estate registration facts or transaction transfer facts. If there are subsequent changes in the status of rights, the system does not directly reuse the contract filing fact, but determines its verification application status as conflicting or insufficiently proven, and triggers cross-source conflict verification or supplementary data acquisition. If the subject identifiers, property locations, and time sequences among the contract filing facts, real estate registration facts, and transaction transfer facts can all be confirmed to be consistent, and there are no invalid or conflicting states, the system can determine the candidate facts as directly applicable.

[0049] In another specific embodiment, the target government service item needs to verify whether the business scope or industry category of the market entity meets the applicable scope of the item. The system obtains the structured text of the business scope from the market entity registration data source and obtains the applicable catalog or industry catalog from the item rule data source. The system splits the business scope text into multiple business scope fact state units and converts the applicable catalog into catalog fact state units or derived strategy inputs.

[0050] When the applicable catalog or catalog mapping relationship changes, the system determines the set of derived facts for the affected business scope based on the version dependency information and only recalculates the affected set. If the mapping relationship between a certain business scope fact status unit and the applicable catalog is clear and the version dependency is not invalid, the system can determine its verification application status as directly applicable; if the catalog version change causes the original mapping relationship to become invalid, the system will determine that the corresponding derived facts need to be recalculated; if there are multiple possible interpretations of the mapping relationship or the proof link is incomplete, the system will determine that the proof is insufficient or generate a review task.

[0051] In the above embodiments, the review results can be further incorporated into subsequent verification strategy parameters. For example, if a reviewer confirms that a certain type of historical contract filing fact cannot independently prove the current housing relationship, the system can adjust the refresh conditions or cross-source conflict verification trigger conditions corresponding to that fact type; if a reviewer confirms that a certain type of directory mapping relationship is invalid in the new version directory, the system can adjust the derived facts and recalculate the trigger parameters. All of the above adjustments must pass constraint invariance verification and must not change the factual determination requirements and logical relationships of the target matter itself.

[0052] This embodiment also provides an adaptive verification system for government affairs based on the factual status of the subject. The system includes a factual status generation module, a subject factual status database module, a factual requirement generation module, a candidate fact matching module, an applicable status determination module, a data processing action execution module, a verification result generation module, a verification execution trajectory module, and a difference location and parameter adjustment module.

[0053] The Fact State Generation module acquires structured data from at least two government data sources and converts it into standardized subject fact state units. Its output is connected to the input of the Subject Fact State Library module. The Subject Fact State Library module stores subject fact state units and provides candidate subject fact state units to the Candidate Fact Matching module. The Fact Requirement Generation module generates fact requirements decoupled from specific government data source interfaces based on verification requests for target government matters. Its output is connected to the input of the Candidate Fact Matching module.

[0054] The candidate fact matching module connects to both the subject fact state database module and the fact-verification requirement generation module, and is used to match candidate subject fact state units corresponding to fact-verification requirements before generating data source access tasks. The applicable state determination module connects to the candidate fact matching module and dynamically determines the verification applicable state based on the objective fact state, fact time information, and current fact-verification requirement of the candidate subject fact state units. The data processing action execution module connects to the applicable state determination module and outputs the fact-verification result when the verification applicable state is directly applicable; when the verification applicable state is not directly applicable, it triggers the corresponding data processing action and writes the processing result to the subject fact state database module.

[0055] The input of the verification result generation module is connected to the data processing action execution module, used to generate automatic verification results based on the subject fact state units that meet the factual verification requirements. The input of the verification execution trajectory module is connected to the candidate fact matching module, the applicable state determination module, the data processing action execution module, and the verification result generation module, used to record the relationships between factual verification requirements, candidate subject fact state units, verification applicable states, data processing actions, data processing results, and automatic verification results. The input of the difference location and parameter adjustment module is connected to the verification execution trajectory module, used to locate the fact verification node causing the difference along the verification execution trajectory when there is a difference between the review result and the automatic verification result, and adjust only the verification strategy parameters affecting the verification applicable state or data processing action.

[0056] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. For those skilled in the art, various changes, modifications, substitutions, or alterations can be made to the above embodiments without departing from the principles and spirit of the invention. The scope of protection of the present invention should be determined by the scope defined in the claims.

Claims

1. An adaptive verification method for government affairs based on the factual status of the subject, characterized in that, include: Obtain structured data from at least two government data sources, convert the structured data into standardized subject fact state units and store them in a subject fact state database; each subject fact state unit includes at least a subject identifier, fact type, fact value, fact time information, data source identifier, and objective fact state. Receive a verification request for the target government matter, and generate a factual requirement decoupled from the specific government data source interface based on the verification request; the factual requirement includes at least the subject to be verified, the type of fact to be proved, the time constraint, and the status of the target fact, but does not include the specific government data source interface identifier and the original field mapping information. Before generating a data source access task for the factual requirement, candidate subject factual state units corresponding to the factual requirement are matched from the subject factual state library; Based on the objective factual status, factual time information, and current factual requirement of the candidate subject factual status unit, the verification applicability status of the candidate subject factual status unit in the current target government affairs is dynamically determined; When the verification application status is the direct application status, the corresponding candidate subject fact status unit is directly reused to generate the factual result, and no data source access task is generated for the factual requirement. When the verification application status is not a directly applicable status, the corresponding data processing action is triggered, and the subject fact status database is updated according to the processing result; The main fact state unit generates an automatic verification result based on the factual requirement that the facts are met, and records the verification execution trajectory. When there is a difference between the review result and the automatic verification result, the fact verification node that caused the difference is located along the verification execution trajectory. Only the verification strategy parameters that affect the verification application status or the data processing action are adjusted, without changing the fact judgment requirements and their logical relationships corresponding to the target government affairs.

2. The method according to claim 1, characterized in that, The main fact state unit includes a source fact state unit and a derived fact state unit; the source fact state unit is directly generated from structured data in the government data source, and the derived fact state unit is calculated and generated from at least one source fact state unit or other derived fact state units, and records at least one version dependency information among the corresponding source fact version, directory version, mapping version or derived strategy version.

3. The method according to claim 1, characterized in that, The objective fact status includes at least one of the following: currently valid, historically valid, pending confirmation, conflicting, invalid, and unknown; the verification applicability status includes at least one of the following: directly applicable, needs to be refreshed, conflicts exist, needs to be recalculated, insufficient proof, and needs to be supplemented.

4. The method according to claim 1, characterized in that, Determining the applicable status of the verification includes making the following judgments in sequence: Does the fact type of the candidate subject fact state unit meet the current factual requirement? Does the fact time information of the candidate subject fact state unit meet the time constraint corresponding to the current factual requirement? Does the fact status unit of the candidate subject conflict with the fact status of the same subject or the same business object in other government data sources? When the candidate subject fact state unit is a derived fact state unit, is its version dependency information invalid? Whether the candidate subject fact state unit can form a complete proof link that meets the current factual proof requirements; The final verification status will be determined based on the above judgment results.

5. The method according to claim 4, characterized in that, When multiple abnormal situations exist, the final verification application status is determined according to the preset status priority; the preset status priority includes: conflict exists above needing to be recalculated, needing to be recalculated above needing to be refreshed, needing to be refreshed above insufficient proof, and insufficient proof above direct application.

6. The method according to claim 4, characterized in that, When determining whether a conflict exists, if there is no unified business object identifier for the subject fact state unit from different government data sources, at least two types of attributes from the subject identifier attribute, location attribute, business object attribute, associated subject attribute, and time attribute used to describe the same object to be verified in different government data sources are standardized, and a cross-source association state is determined based on the standardization result: confirmed association, candidate association, conflict association, or unresolved association.

7. The method according to claim 2, characterized in that, When the source fact version, directory version, mapping version, or derivation strategy version on which the derived fact state unit depends changes, the affected derived fact set is determined based on the reverse index of the version dependency information, and recalculation is performed only on the affected derived fact set, or a recalculation mark is set for the affected derived fact set.

8. The method according to claim 1, characterized in that, The verification execution trajectory includes the relationship between factual requirements, candidate subject factual status units, verification applicable status, data processing actions, data processing results, and automatic verification results. When there is a difference between the review result and the automatic verification result, the automatic verification result is reversed to the corresponding data processing action, verification applicable status, and candidate subject factual status unit according to the verification execution trajectory.

9. The method according to claim 1, characterized in that, Before adjusting the verification strategy parameters, the constraint invariance of the factual determination elements and their logical relationships corresponding to the target government affairs is verified. The constraint invariance verification includes converting the factual determination elements and logical relationships before and after the adjustment into standardized constraint representations, and allowing the adjusted verification strategy parameters to be used for subsequent verification when it is confirmed that the standardized constraint representations before and after the adjustment are consistent.

10. The method according to claim 9, characterized in that, The normalized constraint representation includes fact-determining node and logical relation node; when generating the normalized constraint representation, the fact-determining node is converted to a uniform format, and the multiple child nodes corresponding to the logical relation node that satisfies the exchange relationship are deterministically sorted.

Images