Data verification result analysis method and device, equipment and storage medium

Abstract

The application discloses a data check result analysis method and device, equipment and a storage medium, and relates to the technical field of data analysis. The method comprises the following steps: performing rule matching on a check analysis request input by a user, obtaining a check object identifier and check rule information, and the check rule information comprising rule semantic information and a check execution statement; performing field optimization on the check execution statement according to the rule semantic information, and obtaining an optimized query statement; obtaining abnormal data information based on the optimized query statement and the check object identifier; and generating a check analysis result according to the rule semantic information and the abnormal data information.

Description

Technical Field

[0001] This application relates to the field of data analysis technology, and in particular to a method, apparatus, device and storage medium for parsing data verification results. Background Technology

[0002] Regulatory reporting requires high data quality, so a lot of verification is performed on each reporting product. These verifications strongly guarantee the accuracy of the reported data, but at the same time, they also represent a lot of maintenance work.

[0003] However, when encountering validation errors, the responsible business personnel in the validation department need to coordinate with the IT (Information Technology) personnel, who then query the detailed data item by item and provide the specific reason for the data error. When facing complex validations, such as those involving cross-fields, cross-tables, cross-periods, or cross-product categories, the existing data validation result parsing solution suffers from excessive time consumption.

[0004] Therefore, how to improve the efficiency of verification exception handling has become an urgent problem to be solved. Summary of the Invention

[0005] The main purpose of this application is to provide a data verification result parsing method, apparatus, device and storage medium, aiming to solve the technical problem of how to improve the efficiency of verification anomaly handling.

[0006] To achieve the above objectives, this application proposes a data verification result parsing method, which includes: The user-input verification parsing request is matched with rules to obtain the verification object identifier and verification rule information, wherein the verification rule information includes rule semantic information and verification execution statements; Based on the semantic information of the rules, the validation execution statement is optimized to obtain an optimized query statement; Abnormal data information is obtained based on the optimized query statement and the verification object identifier; A verification and parsing result is generated based on the semantic information of the rules and the abnormal data information.

[0007] Furthermore, to achieve the above objectives, this application also proposes a data verification result parsing device, which includes: The rule query module is used to perform rule matching on the user-input verification parsing request to obtain the verification object identifier and verification rule information, wherein the verification rule information includes rule semantic information and verification execution statement; The statement optimization module is used to optimize the fields of the validation execution statement based on the rule semantic information to obtain an optimized query statement; An anomaly detection module is used to obtain abnormal data information based on the optimized query statement and the verification object identifier; The data parsing module is used to generate verification and parsing results based on the rule semantic information and the abnormal data information.

[0008] In addition, to achieve the above objectives, this application also proposes a data verification result parsing device, which includes: a memory, a processor, and a data verification result parsing program stored in the memory and executable on the processor. The data verification result parsing program is configured to implement the steps of the data verification result parsing method described above.

[0009] In addition, to achieve the above objectives, this application also provides a storage medium storing a program that implements the data verification result parsing method, wherein the program that implements the data verification result parsing method is executed by a processor to implement the steps of the data verification result parsing method as described above.

[0010] This application provides a method, apparatus, device, and storage medium for parsing data verification results. The method includes: performing rule matching on a user-input verification parsing request to obtain a verification object identifier and verification rule information, wherein the verification rule information includes rule semantic information and a verification execution statement; optimizing the fields of the verification execution statement based on the rule semantic information to obtain an optimized query statement; obtaining abnormal data information based on the optimized query statement and the verification object identifier; and generating a verification parsing result based on the rule semantic information and the abnormal data information.

[0011] This application optimizes the fields of a pre-configured validation execution statement to obtain an optimized query statement, thereby accurately locating abnormal data based on the optimized query statement, reducing the need for manual intervention and improving the efficiency of validation anomaly handling. Attached Figure Description

[0012] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0013] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0014] Figure 1 This is a flowchart illustrating the first embodiment of the data verification result parsing method of this application; Figure 2 This is a flowchart illustrating the second embodiment of the data verification result parsing method of this application; Figure 3 This is a flowchart illustrating the third embodiment of the data verification result parsing method of this application; Figure 4 This is a simplified schematic diagram of the data verification result parsing method of this application; Figure 5 This is a schematic diagram of the module structure of the data verification result parsing device according to an embodiment of this application; Figure 6 This is a schematic diagram of the device structure of the hardware operating environment involved in the data verification result parsing method in this application embodiment.

[0015] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0016] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application. Furthermore, all actions involving the acquisition of signals, information, or data in this application are performed in accordance with the relevant data protection regulations of the country where the application is located and with authorization from the owner of the corresponding device.

[0017] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0018] It should be noted that the execution subject in this embodiment can be a data verification result parsing system, or a computing service device with data processing, network communication, and program execution functions, such as a tablet computer, personal computer, or mobile phone, or a data verification result parsing device capable of performing the above functions. This embodiment does not specifically limit it in this way. The following uses a data verification result parsing device (hereinafter referred to as the parsing device) as the execution subject as an example to describe this embodiment and the following embodiments.

[0019] Based on this, this application provides a data verification result parsing method. This method is designed for financial regulatory reporting data verification scenarios. It can automate the entire process of rule matching, SQL (Structured Query Language) optimization, data querying, and intelligent parsing, transforming the technical results of verification errors into error reasons and details that business personnel can directly understand.

[0020] Therefore, refer to Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the data verification result parsing method of this application.

[0021] In this embodiment, the data verification result parsing method includes steps S10~S40: Step S10: Perform rule matching on the user-input verification parsing request to obtain the verification object identifier and verification rule information, wherein the verification rule information includes rule semantic information and verification execution statement; It's easy to understand that the aforementioned verification parsing request can refer to a natural language query request submitted by users (on the verification display platform) such as the responsible business personnel. Users can use this verification parsing request to inquire about the reason why a verification object under a specific reporting period or a specific business data block triggered an error. For example, the content of this verification parsing request could be "Why did verification project A report an error in data block B in the fourth quarter of xx year?" In this case, the aforementioned rule matching can refer to the process of extracting key parameters from the user-input verification parsing request and matching them with corresponding verification rules through semantic recognition and knowledge base retrieval.

[0022] It is understood that the aforementioned verification object identifier can be a verification object, such as the data block identifier corresponding to a certain verification rule or a certain verification item, such as a data block ID (Identity) or a data block number. In this embodiment, the verification object identifier can be a unique identifier that distinguishes the data scope of different business lines / business personnel under the same data table and the same data period. That is, this embodiment can realize data permission isolation based on the verification object identifier, ensuring that business personnel can only view the data within their own scope of responsibility.

[0023] It is important to understand that the verification rule information can be the full set of structured verification rules stored in the rule knowledge base. This is the core basis for subsequent verification execution and cause analysis in this embodiment, and it can contain two core types of content: rule semantic information and verification execution statements. In this case, after receiving the verification parsing request submitted by the user, the parsing device can first extract the core parameters using natural language processing technology, and then access the rule knowledge base using the parameters as search conditions to match and obtain the corresponding verification rule information and verification object identifier.

[0024] The rule semantic information can include a verification rule description and verification fields. The verification rule description can be a verification business logic, compliance requirements, and applicable scenarios expressed in natural language, representing the standard for judging whether the data in the verification parsing request is compliant at the business level, such as "If the real-time amount of verification item A is less than 0, please report it"; while the verification fields can correspond to specific verification objects, such as field A corresponding to verification item A.

[0025] The aforementioned validation execution statement, also known as a validation SQL statement, refers to a structured query statement that implements validation logic and filters abnormal data in the database. This validation execution statement can fully encompass table joins, filtering conditions, and calculation logic, serving as the direct carrier of data-level validation execution. For example, the execution logic of this validation execution statement can be expressed as: check if the real-time value of field A is less than 0; if so, report it.

[0026] Understandably, existing manual rule lookup methods require manually entering multiple structured parameters such as rule number, period selection, and data block selection to search for and query verification issues in a large number of rule documents. This is extremely inefficient and prone to selecting the wrong rule version or an invalid rule, leading to errors in all subsequent analyses. Furthermore, different business personnel have inconsistent understandings of the rules, and manual rule matching is prone to bias, affecting the accuracy of the parsing results. Therefore, in a feasible implementation, in this embodiment, step S10 may include steps A1~A2: Step A1: Perform semantic parsing on the user-inputted verification parsing request to obtain verification question elements, which include: data period information, verification object identifier, and verification rule identifier; It is easy to understand that in the semantic parsing process described above, the parsing device can pre-train a named entity recognition model based on Natural Language Processing (NLP) technology, specifically tailored to the expression habits of the verification question scenario. Then, the named entity recognition model identifies and extracts structured key parameters—the aforementioned verification question elements—from the unstructured natural language request text contained in the verification parsing request. These verification question elements can be three types of core structured parameters constituting a complete verification parsing request. Specifically, they may include data period information, verification object identifier (data block ID), and verification rule identifier.

[0027] The data period information can be the statistical reporting cycle corresponding to the verification data, which can be used as a time filter condition for subsequent data queries. It needs to be uniformly converted into a standard date format after parsing (the end date of the quarter is taken by default for quarters and the end date of the month is taken by default for months). For example, "2025.12.31" and "2025 / 12 / 31" can both be converted into "2025-12-31", and the fourth quarter of 2025 can also be converted into "2025-12-31".

[0028] The aforementioned verification rule identifier can be a unique identifier for a verification rule in the subsequent rule knowledge base. It can correspond to the primary key for retrieval in the rule knowledge base and is usually composed of the business code (product code / business line) corresponding to the business of the verification data plus the serial number, such as "M1_01".

[0029] For example, after receiving the natural language verification request text input by the user, the parsing device can first perform basic text cleaning, removing redundant spaces, special symbols, and meaningless interjections. Then, it loads a named entity recognition model fine-tuned for regulatory reporting verification scenarios, configuring recognition rules for the three types of entities corresponding to the aforementioned verification issue elements. For example, for time entities, it can recognize various time expressions such as "XXXX year X quarter", "XXXX year X month", "XXXXQ4", and "XXXX year end", and uniformly convert them into data period information in the YYYY-MM-DD standard format; for rule number entities, it can identify verification rule numbers conforming to the format "product code_serial number" through regular expression matching + dictionary matching, using these as verification rule identifiers; for data block entities, it can identify data block IDs conforming to the format "M business code-submodule code", using these as verification object identifiers.

[0030] Furthermore, the named entity recognition model can output recognition results and confidence scores for three types of entities. Entities with a confidence score ≥ 0.9 can be directly identified as valid elements. If there are elements with insufficient confidence or missing recognition, the parsing device can ask the user follow-up questions to confirm the corresponding parameter information. Therefore, this embodiment supports natural language questions, meaning users do not need to remember technical parameters such as rule number formats and data block codes. They can initiate the verification question parsing process using everyday business expressions, and the parsing device can automatically guide the user to input missing information, greatly improving usability.

[0031] Finally, after all three types of entity information have been confirmed as trustworthy, the Named Entity Recognition model assembles the qualified entity information into structured verification question elements for output.

[0032] Step A2: Based on the data period information and the verification rule identifier, query the rule knowledge base to obtain the verification rule information.

[0033] It's important to understand that the rule knowledge base can be a pre-built structured verification rule repository stored internally by the parsing device. This rule knowledge base can store the full configuration information of all regulatory reporting verification rules and support quick retrieval based on the aforementioned verification rule identifiers. In this case, all verification rules in the rule knowledge base can be imported from Excel, sliced, and processed to generate HTML format key-value pair data storage, including rule text descriptions, verification SQL, verification fields, effective period, applicable business, and other fields.

[0034] At this point, the parsing device can use the extracted verification rule identifier and data period as search conditions, perform an exact matching query in the rule knowledge base, filter out the corresponding rules with effective status, and parse and extract structured information to obtain the verification rule information.

[0035] For example, the parsing device can verify that the rule identifier is the retrieval primary key, construct an SQL query statement, and access the rule configuration table in the rule knowledge base. Simultaneously, the parsing device can also append an effective period verification to the query conditions: "Data Period BETWEEN Rule Effective Start Date AND Rule Expiration Date," ensuring a match with a valid rule for the current period. Then, based on the statement with the appended period verification, the device executes a rule knowledge base query to retrieve the matched rule records, which are stored as a key-value pair table in HTML format.

[0036] Next, the parsing device can use regular expressions and HTML parsing tools to extract the core content of the corresponding validation rule information, such as "rule description," "validation SQL," "validation fields," and "scope of application," from the table. This extracted core content is then assembled into structured validation rule information, including rule semantics and validation execution statements. Furthermore, if the parsing device fails to find a matching valid rule in the rule knowledge base, it can return an empty value and provide the user with a message similar to "No corresponding valid validation rule found. Please check if the rule number and period are correct."

[0037] In this implementation, all rules are centrally maintained and updated in a knowledge base, effectively avoiding version inconsistencies caused by rules being distributed among different personnel. Furthermore, this embodiment utilizes a dual verification method—precise matching by rule number and validity period—to ensure automatic matching of currently valid and correct rules from the rule knowledge base, avoiding analytical biases caused by version errors and improving rule matching accuracy. Therefore, this embodiment can automatically query a standardized knowledge base to ensure the accuracy and effectiveness of rule matching, lowering the user's barrier to entry and improving the efficiency and accuracy of rule matching.

[0038] Step S20: Optimize the fields of the verification execution statement according to the rule semantic information to obtain an optimized query statement; Understandably, the aforementioned field optimization can refer to the pruning and simplification of the returned fields of the validation execution statement based on the semantic logic and related fields of the validation rules, without changing the original validation and filtering logic, table structure, and relationships.

[0039] In this case, the optimized query statement can be the final SQL query statement after field optimization and permission condition concatenation. Its corresponding data range can include only the primary key field and the validation-related fields. Therefore, in this embodiment, the optimized query statement can return only the primary key field and the validation-related fields, thus enabling accurate and efficient querying of abnormal data within a specified range.

[0040] For example, the parsing device can parse the syntax structure of the validation execution statement, identifying the main table, related tables, table aliases, and WHERE filter conditions in the statement. Then, combined with the rule semantic information, it extracts all business fields involved in the WHERE condition as preliminary fields. After optimizing the preliminary fields, it combines the FROM clause, JOIN relationship, WHERE filter conditions, GROUP BY grouping logic, and ORDER BY sorting logic of the original statement to generate a modified SQL statement as the optimized query statement, ensuring that the validation logic of the optimized query statement is completely consistent with the original statement.

[0041] Step S30: Obtain abnormal data information based on the optimized query statement and the verification object identifier; It is easy to understand that the aforementioned abnormal data information can be a dataset that does not conform to the validation rules and is returned after the optimized query statement is executed in the database, serving as the data basis for subsequent error analysis. Furthermore, in this embodiment, the abnormal data information can also be presented as full detailed data or grouped statistical data, depending on the data volume. In this case, the parsing device can supplement the data permission filtering conditions based on the optimized query statement and submit it to the database for execution, obtain the query results, and perform adaptation processing to obtain structured abnormal data information.

[0042] Step S40: Generate a verification and parsing result based on the rule semantic information and the abnormal data information.

[0043] It should be understood that the above-mentioned verification and parsing results can refer to using the semantic information of the rules as the judgment standard, combining the actual values ​​and distribution of abnormal data to perform attribution analysis, and generating natural language parsing conclusions that business personnel can directly understand. The verification and parsing results may include the total number of errors, explanation of the error reasons, error details / statistical results, and rectification reference directions, etc.

[0044] For example, the parsing device can load a preset verification and parsing prompt word template, fill in the rule semantic information, verification fields, and abnormal data information into the corresponding positions of the template, and then input the completed prompt words into the large language model. The constraint model outputs according to the structure of "rule requirement description - overall error situation - detailed / statistical examples - reason summary". Then, the large language model combines the compliance requirements of the rules, analyzes the specific reasons why the field values ​​of the abnormal data do not conform to the rules, and generates fluent, professional, and business-terminologically appropriate natural language parsing text.

[0045] Furthermore, the parsing device can also perform basic compliance checks on the model output content to ensure that the description is consistent with the actual data and that there is no fabricated content. Then, it returns the final verification and parsing result to the verification and display platform.

[0046] In summary, this embodiment can optimize and query based on standardized original verification SQL and actual conditions, so that the verification logic of the optimized query statement remains unchanged. Combined with knowledge base rule constraints, it can effectively ensure that the parsing accuracy is much higher than the average accuracy of existing manual analysis.

[0047] Meanwhile, this embodiment can realize a fully automated closed loop, from user inquiry to result output without human intervention, forming an automated processing closed loop of "problem input - intelligent analysis - solution output". Through this automated process, the efficiency of handling a single problem is improved, compressing the existing manual hour-level verification and error parsing process to the second level. A single reporting cycle can save more than 6 person-days of IT operation and maintenance workload, effectively alleviating the support pressure during peak reporting periods.

[0048] Furthermore, in this embodiment, business personnel do not need to understand SQL syntax and underlying data logic. They can obtain clear error reasons and details by asking questions in natural language, which significantly improves their ability to handle verification issues independently. This allows business personnel to solve routine verification problems on their own, thereby lowering the threshold for business processing.

[0049] This embodiment provides a data validation result parsing method, which includes: performing rule matching on a user-input validation parsing request to obtain a validation object identifier and validation rule information, wherein the validation rule information includes rule semantic information and a validation execution statement; parsing key fields of the rule semantic information to obtain validation-related field information; parsing the execution subject of the validation execution statement to obtain query subject information; adjusting the return field range of the validation execution statement based on the query subject information and validation-related field information to obtain an optimized query statement; obtaining abnormal data information based on the optimized query statement and the validation object identifier; and generating a validation parsing result based on the rule semantic information and the abnormal data information.

[0050] This embodiment optimizes the pre-configured verification execution statement by performing field optimization to obtain an optimized query statement. Based on the optimized query statement, abnormal data can be accurately located, thereby reducing the need for manual intervention and improving the efficiency of verification anomaly handling.

[0051] Based on the first embodiment of this application, in the second embodiment of this application, the same or similar content as the first embodiment described above can be referred to the above description, and will not be repeated hereafter.

[0052] It's easy to understand that optimizing SQL validation in existing technologies often leads to two core problems: First, the risk of rewriting accuracy, meaning that during optimization, the inner SELECT statement, the join logic, or the filtering conditions can be arbitrarily modified, causing the validation logic to be broken. This results in erroneous data retrieved that is inconsistent with the official validation results, leading to a loss of parsing accuracy and even syntax errors that cause SQL execution to fail. Second, unreasonable field pruning, meaning that during statement optimization, either a large number of irrelevant fields are retained, failing to achieve the effect of simplifying data and improving efficiency, or primary keys or core validation fields are omitted, making it impossible for business personnel to locate specific erroneous data or analyze the cause of the error.

[0053] To address this issue, the field optimization process proposed in this embodiment uses rule semantic information as a basis to identify related fields involved in the validation logic. While preserving the integrity of the original validation logic, the range of fields returned by the SELECT statement in the validation execution statement is adjusted to generate a simplified query statement. At this point, based on the first embodiment, please refer to... Figure 2 , Figure 2 This is a flowchart illustrating the second embodiment of the data verification result parsing method of this application. In this embodiment, step B0 includes steps B1 to B3: Step B1: Parse the key fields of the rule semantic information and the verification execution statement to obtain the verification-related field information; It is easy to understand that the above-mentioned key field parsing is a preparatory step for subsequent field optimization in this embodiment. During this process, the parsing device can perform field extraction through two channels. First, it can map the rule semantic information to keywords using a preset business field dictionary to obtain the first set of candidate fields. Then, by parsing the syntax of the verification execution statement, it can determine the business fields involved in the verification execution statement as the second set of candidate fields. Next, it can take the union of the two sets of candidate fields and remove duplicates, and then compare them with the field metadata of the corresponding data table to filter out invalid fields that are incorrectly identified, so as to ensure that all fields actually exist in the corresponding table. Finally, the parsing device can assemble the fields that pass the verification into structured verification-related field information and mark the table alias corresponding to each field for subsequent concatenation. This verification-related field information can correspond to the fields that determine whether the data triggers a verification error, and is usually the field that relevant business personnel need to focus on when investigating the cause of the error during error analysis.

[0054] At this point, the parsing device performs natural language semantic recognition on the natural language description of the rule semantic information, identifies the business keywords involved, and then matches the identified business keywords with the preset business field dictionary to obtain the business fields directly constrained by the verification logic corresponding to the rule semantic information, namely the first group of candidate fields mentioned above.

[0055] Meanwhile, the parsing device can parse the syntax tree of the verification execution statement through the SQL syntax parser, extract all business fields involved in the WHERE clause of the verification execution statement or the verification calculation logic, and obtain the above-mentioned second set of candidate fields.

[0056] Step B2: Parse the execution subject of the verification execution statement to obtain the query subject information; Understandably, the aforementioned execution subject parsing can refer to using SQL syntax parsing technology to break down the structure of the validation execution statement, identify the main table, related tables, and table aliases involved in the statement, and then match them with preset data table metadata configurations to obtain the corresponding table's primary key field. Correspondingly, the aforementioned query subject information can refer to the primary key field configurations of each table involved in the validation execution statement.

[0057] At this point, the parsing device first uses an SQL syntax parser to break down the FROM and JOIN clauses of the validation execution statement, identifying all table names and corresponding table aliases involved in the statement, and determining the main table of the statement (i.e., the first table after the outermost FROM, which is usually the main table of the erroneous data). Then, for each table, it accesses the preset table metadata configuration table and matches the unique primary key field corresponding to the table based on the main table name (e.g., "BUS_ID" for business tables and "CUST_ID" for customer tables). If the validation involves multiple table joins, the primary key field of each table is obtained separately, with the primary key of the main table being retained as the core primary key. This primary key field can be used to accurately locate a single erroneous data record and is a field that must be retained in the statement's returned results. Next, the parsing device assembles the table names, table aliases, and primary key fields into structured query body information, completing the above-mentioned execution body parsing.

[0058] Step B3: Based on the query subject information and the verification related field information, adjust the range of returned fields of the verification execution statement to obtain an optimized query statement.

[0059] It is easy to understand that the aforementioned adjustment of the returned field range can be considered the core operation of this embodiment in optimizing fields by focusing on the problem fields. Specifically, this embodiment can modify the field list of the SELECT clause of the SQL statement, removing all fields unrelated to the validation logic and retaining only the primary key and validation-related fields to adjust the range of returned fields. Therefore, this embodiment can ensure that the returned fields of the optimized query statement retain only "the primary key necessary for locating data + the validation-related fields necessary for analyzing the cause," which minimizes data transmission volume, reduces database and large model resource consumption, and meets the needs of business personnel for troubleshooting and error reporting, avoiding information omissions.

[0060] At this point, the parsing device can assemble the "primary key field + all validation related fields" into a new SELECT field list based on the table alias in the query subject information. Then, strictly following the security constraint of "only modifying the outermost SELECT", it replaces the field content after the first (outermost) SELECT of the validation execution statement with the new SELECT field list. This ensures that the optimized execution statement after replacement can still completely retain all the logic of the original validation execution statement, such as table associations, filtering conditions, grouping and sorting.

[0061] For example, the parsing device can query the main table alias in the main information as a prefix, and then concatenate "main table primary key field + all validation related fields" sequentially, with the fields separated by commas, to form a new SELECT field list. For instance, if the main table alias is "a", the primary key is "BUS_ID", and the validation related field is "FXJBL_RMB_SSD", then the assembled result after concatenation can be "a.BUS_ID,a.FXJBL_RMB_SSD".

[0062] Then, the parsing device can locate and verify the first (outermost) SELECT clause of the execution statement, and convert the original full-field query (such as "SELECT a") to a SELECT clause. "SELECT" The entire list of fields is replaced with the one concatenated in the previous step, removing irrelevant fields. At the same time, all other elements of the original SQL statement, including the FROM clause, all JOIN relationships, WHERE filtering conditions, GROUP BY grouping, ORDER BY sorting, subqueries, and window functions, remain completely unchanged, ensuring that the validation logic of the rewritten SQL statement after field optimization is entirely consistent with the original statement.

[0063] Furthermore, the parsing device can also perform syntax verification on the rewritten SQL statement to confirm that there are no missing fields, alias conflicts, or syntax errors. After the verification is passed, the output is the final optimized query statement.

[0064] Therefore, this embodiment strictly limits the modification to only the return fields of the outermost SELECT statement, so that the optimized query statement completely retains all the validation logic of the original SQL, thereby eliminating logical deviations caused by rewriting and ensuring that the optimized SQL query result is completely consistent with the original validation SQL.

[0065] Furthermore, this embodiment can reduce the number of fields returned by the optimized query statement, thereby reducing database scanning and transmission overhead and improving SQL execution speed, especially for wide table scenarios where the optimization effect is significant. Meanwhile, in manual verification scenarios, business personnel can quickly locate the value of problematic fields by using the returned results containing only the primary key and verification-related fields, avoiding searching through a large number of irrelevant fields and improving problem location efficiency. In intelligent verification scenarios, it can reduce the token consumption and system cache usage of large models by reducing the amount of returned data, avoiding the risk of memory overflow under large data volumes, improving overall system stability, and reducing system resource consumption.

[0066] In summary, this embodiment proposes a standardized SQL query optimization process: "Extracting validation fields → Parsing query subject metadata → Securely rewriting the SELECT range." Understandably, this optimization process has two core objectives: first, to ensure rewriting accuracy by parsing fields and table structures before rewriting, thus preventing issues such as alias conflicts, missing fields, and tampering with validation logic, ensuring that the optimized SQL query result is completely consistent with the original validation SQL; second, to achieve precise tailoring by retaining only the primary key and validation-related fields necessary for troubleshooting, minimizing the amount of returned data and reducing system resource consumption without affecting error analysis. This three-step process logic can be fully automated by SQL parsing tools without manual intervention, effectively adapting to validation SQL statements of varying complexity, and possessing versatility and feasibility.

[0067] It is understandable that existing query processes pose a security risk of unauthorized access to sensitive data tables and retrieval of data outside of authorized scope, leading to the leakage of sensitive financial data. Therefore, in a feasible implementation, in this embodiment, step S30 includes steps S31-S33: Step S31: Based on the verification object identifier, execute the optimized query statement with restricted access permissions to obtain the initial query result; It's easy to understand that restricting access permissions as described above refers to limiting the range of data queried during statement execution by appending a data block ID filter condition to the query SQL, ensuring that users can only access abnormal data within the data block they are responsible for. In this case, the parsing device can append a verification object identifier as a data block permission filter condition to the optimized query statement before submitting it to the database for execution, ensuring that the optimized query statement can only retrieve abnormal data within the specified permission range. Therefore, the initial query result described above can be the complete list of raw abnormal data obtained after executing the optimized query statement with the appended access permission condition in the database.

[0068] At this point, the permission filtering condition "AND BLK_ID='verification object identifier'" can be added to the WHERE condition of the optimized query statement to achieve permission isolation for the data range. Then, the optimized SQL statement with the verification object identifier is submitted to the database execution engine for permission verification of the whitelist. Only after the verification passes can the SQL query be executed; otherwise, the execution of the optimized query statement is rejected. Therefore, this embodiment can achieve fine-grained permission isolation through verification object identifier, ensuring data access compliance and eliminating the risk of unauthorized access. In a feasible implementation, step S31 in this embodiment includes steps S311 to S312: Step S311: Determine whether the document to be accessed corresponding to the optimized query statement belongs to the range of accessible documents corresponding to the verification object identifier; It's easy to understand that the document to be accessed, also known as the data table to be accessed, can refer to all business data tables involved in the FROM and JOIN clauses of the optimized query statement, serving as the data source for this query. In this case, the accessible document range can also be called a whitelist, meaning it can be a set of regulatory reporting business data tables pre-configured by the administrator and allowed access by the intelligent parsing system. The parser extracts all table names from the optimized query statement and compares them one by one with the accessible whitelist to verify that all tables are within the allowed access range. Only tables within the accessible document range corresponding to the optimized query statement can be queried and executed by the optimized query statement, thus implementing a security control mechanism at the data source level.

[0069] For example, after receiving the optimized query statement and the verification object identifier (data block ID), the WHERE clause of the optimized query statement can be parsed to confirm whether the BLK_ID filtering condition already exists; if it does not exist, the parsing device can append the filtering condition "AND BLK_ID=[verification object identifier]" to the end of the WHERE clause.

[0070] Then, the parsing device uses its built-in SQL syntax parser to deeply analyze the structure of the optimized query statement, extracting all table names appearing in the FROM clause and all JOIN clauses, and removing the schema prefix to obtain a list of pure table names. The parsing device can then read the system's preset whitelist of accessible tables, which may include business table data related to regulatory reporting. From this list, the data identifier of the aforementioned appended optimized query statement can be obtained, i.e., the range of accessible documents corresponding to the verification object identifier.

[0071] Then, the parsing device compares the names of all data tables extracted for the optimized query statement with the accessible document range one by one. If all data tables are within the accessible document range, the permission verification is considered successful, and the query is allowed to be executed. If any data table is outside the accessible document range, the permission verification is considered unsuccessful, and the optimized query statement is rejected. Furthermore, if the permission verification fails, the query process can be terminated immediately, and a message "No access permission to the corresponding data table, please confirm the query range" will be returned to the querying user. Simultaneously, an abnormal access log will be recorded for subsequent security auditing. In this way, this embodiment ensures that the query range of the optimized query statement is limited to the specified data block.

[0072] Step S312: If it is detected that the document to be accessed belongs to the scope of accessible documents, the output restriction execution result corresponding to the optimized query statement is used as the initial query result.

[0073] Understandably, existing statement execution processes may involve excessive database CPU, I / O, and connection resources being consumed by queries with extremely large data volumes, potentially impacting the normal operation of other core reporting services. In this case, the aforementioned output limit can refer to the output result determined by applying an upper limit constraint to the maximum number of data rows returned after performing an SQL query on the optimized query statement. That is, this embodiment can append a maximum return row limit to the validated SQL statement before submitting it to the database for execution, ensuring that the returned data volume does not exceed a safe threshold.

[0074] For example, after the data access permission verification based on the accessible document range passes, the parsing device can further read the system's preset maximum output record limit (the default configuration is 1000 records, which can be adjusted according to database performance; this embodiment does not impose this limit). Then, it appends the "LIMIT [maximum output record count]" restriction statement to the end of the optimized query statement to constrain the maximum number of records returned by the database. The SQL statement with the output limit appended is then submitted to the database execution engine. At this point, after the database executes the optimized query statement, it can return abnormal data details that do not exceed the maximum record limit. The parsing device receives the database's query return result as the initial query result and enters the subsequent processing flow.

[0075] Therefore, this embodiment clearly defines a dual security protection mechanism of "table list verification + output number limit" to avoid the risks of illegal table access and excessive query from the two dimensions of data source and output volume, ensuring that data access behavior is controllable and compliant throughout the process, and guaranteeing the safe and stable operation of the system in the production environment.

[0076] Step S32: Obtain the data volume characteristics of the initial query results, and determine the current output strategy based on the data volume characteristics; Understandably, the data volume characteristic can be the total number of data entries in the initial query results, which can serve as the core quantitative basis for subsequently determining which output strategy to adopt. That is, in this embodiment, the optimal output strategy can be automatically switched and selected based on the size of the data volume queried in real time. At this time, the parsing device can pre-set different abnormal data output processing schemes for different data volumes, which can include two types: detailed output strategy and grouped statistical output strategy. That is, the current output strategy mentioned above can be a detailed output strategy or a grouped statistical output strategy.

[0077] The detailed output strategy is suitable when the number of erroneous data entries is small. It can directly return the full details of the abnormal data, including the primary key and validation field values ​​for each data entry. The parsing device can accurately locate a single erroneous data entry based on this detailed output strategy.

[0078] The aforementioned grouping and statistical output strategy is suitable when there are a large number of erroneous data entries. It groups data according to the values ​​of the verification-related fields and then counts the number of erroneous entries in each group. The parsing device can use this grouping and statistical output strategy to understand the overall distribution pattern of errors and avoid information overload caused by a large number of details.

[0079] Then, the parsing device can statistically analyze the data volume characteristic value of the initial query results and compare it with a preset threshold, selecting the corresponding output strategy based on the comparison result. For example, the parsing device can receive the query result set returned by the database, count the total number of data entries as the data volume characteristic value, and then select the current output strategy based on a preset threshold (default configuration is 5 entries, adjustable by the user according to business needs). At this point, if the total number of entries is less than or equal to the preset threshold, it is determined to be a small data volume scenario, and a detailed output strategy can be executed; if the total number of entries is greater than the preset threshold, the number of erroneous entries can be counted by grouping according to the validation correlation fields, and the grouped statistical output strategy can be used as the current output strategy.

[0080] Step S33: Process the initial query result based on the current output strategy to obtain abnormal data information.

[0081] Finally, the parsing device can assemble the processed results into structured exception data information, including the total number of errors and detailed / grouped statistical data, according to the selected current output strategy. For example, if the current output strategy is a detailed output strategy, the parsing device can retain all detailed data from the initial query results, with each data entry containing the values ​​of the primary key field and all validation-related fields; then, it sorts the initial query results in ascending order by the primary key and assembles them into a detailed data list. Ultimately, the exception data information may include: the total number of errors, data type identifier (details), and the detailed data list.

[0082] If a grouped statistical output strategy is adopted, the parsing device can extract all validation correlation fields except the primary key as grouping dimensions; then, it can group and aggregate the initial query results according to the grouping dimensions, counting the number of erroneous data entries under each group; finally, it can sort the grouped results in descending order of the number of erroneous entries, so that users can first view the value types with the highest error count. Ultimately, the abnormal data information can include: total number of erroneous entries, data type identifier (grouped statistics), and a list of grouped statistical results. In this case, the parsing device can use grouped statistics to reduce the amount of returned data in large data volume scenarios, reduce database transmission, system caching, and large model token consumption, and improve the system's concurrency capacity and stability.

[0083] Therefore, this embodiment can achieve a balance between performance and user experience by using an adaptive output strategy to ensure the ability to locate detailed problems with small amounts of data while avoiding resource consumption and information overload with large amounts of data.

[0084] Understandably, code value verification is a common type of verification rule. Its verification logic checks whether the value of a certain field is within a specified standard code value range. For example, the country code must be in the central bank's standard country code table, and the currency code must be in the standard currency table. Existing solutions for code value verification errors involve steps where business personnel are unaware of the correct code value range and need to additionally consult the code value table or IT personnel, resulting in low processing efficiency. Therefore, in a feasible implementation, in this embodiment, step S40 includes steps S41-S43: Step S41: Detect whether there is a code value verification code in the optimized query statement; Step S42, if yes, then query the code value configuration table based on the code value verification encoding to obtain code value example data; Step S43: Generate a verification and parsing result based on the code value example data, the rule semantic information, and the abnormal data information.

[0085] It's easy to understand that the code value verification code, also known as the code value type number, is a unique identifier for each type of code value in the code value configuration table. It typically follows a format starting with a fixed letter followed by different numbers, such as country code F0001, currency code F0002, and business type code F0012. In this case, the parsing device can use regular expressions to match the text content of the verification execution statement, and combine the matching results with semantic analysis to identify whether the statement contains a code value table reference and the corresponding code value type number, thereby achieving the detection of the aforementioned code value verification code.

[0086] For example, the parsing device can use preset regular expression matching rules to search for code value type identifiers in the optimized query SQL text, such as "CODE_TYP='FXXXX'" or "code value type='FXXXX'". Simultaneously, it can combine rule semantic information to determine whether the rule description contains keywords such as "code value," "encoding," "dictionary," or "reference," assisting in verifying whether it is a code value-based verification. If a valid code value verification code is matched, it can be extracted as the basis for subsequent queries, indicating the existence of code value verification, and proceeding to the next step. If no code value verification code is matched, it can be determined as a non-code value-based verification, skipping the code value query process and directly entering the regular parsing stage.

[0087] Understandably, the aforementioned code value configuration table can be a standardized code value lookup table pre-stored in a knowledge base. This table can be divided into tables according to code value type, and includes fields such as code value encoding, code value name, validity, and effective date. In this case, the code value example data can be several compliant code values ​​and their corresponding names randomly selected from the valid code values ​​of the corresponding type. This example data can provide business personnel with an intuitive and correct reference, thereby assisting in data rectification and checking data correctness.

[0088] For example, the parsing device can use the extracted code value verification code as a search condition to access the code value configuration master table in the knowledge base, query the code value detail table name corresponding to this type, and access the corresponding code value detail table to query all code value records with "validity='valid'", obtaining the code value code and corresponding name. Then, it randomly selects 3-5 records from the valid code value list as code value examples, and assembles the extracted code value codes and names into structured code value example data.

[0089] Then, the parsing device can supplement the verification and parsing results with compliant code value example data, or combine the error data to explain the reason for the code value error, and supplement the data to provide business personnel with rectification reference.

[0090] Therefore, this embodiment can also achieve one-stop parsing of code value verification. The system automatically identifies the code value type verification and matches the corresponding code value table, eliminating the need for manual selection by the user. It is adaptable to all code value verification scenarios with no additional usage cost. Furthermore, this embodiment can directly provide correct code value examples in addition to the error reason, helping business personnel quickly complete data rectification without repeated consultation and confirmation. This improves both the quality of reported data and the efficiency of verification problem processing.

[0091] Based on the first and / or second embodiments of this application, in the third embodiment of this application, the content that is the same as or similar to that in embodiments one and two above can be referred to the above description, and will not be repeated hereafter. Based on this, please refer to... Figure 3The data verification result parsing method further includes: In one feasible implementation, the method further includes steps S11-S12: Step S11: When the current parsing mode is intelligent parsing mode, the user-inputted verification parsing request is input into the preset parsing model for step-by-step parsing to obtain the initial parsing content. The initial parsing content includes at least one of the following: verification problem elements, verification rule information, optimized query statement, code value example data, and abnormal data information. In simple terms, the aforementioned intelligent parsing mode refers to a parsing mode that simulates and demonstrates the complete thought process and operation of an IT expert in handling verification issues, based on the reasoning and tool invocation capabilities of a large language model, and displays the final data verification result. This mode can stream the execution status and results of each step, making the process transparent and traceable. This helps business personnel become familiar with the specific verification error handling process. Given sufficient time, users can independently learn from the specific verification process output by the model in this mode, thereby enabling them to independently discover and resolve problematic data, achieving the technical effect of empowering users' professional capabilities.

[0092] Therefore, the aforementioned preset parsing model can be a verification anomaly intelligent parsing agent (intelligent agent), which is a dedicated intelligent agent built based on a large language model. It can be pre-injected with IT expert personas, the standardized processing flow of the above embodiments one and two, a dedicated toolchain and business constraint rules, and can autonomously complete the entire process of verification problem parsing without the need for manual scheduling tools.

[0093] It is important to understand that in the aforementioned step-by-step parsing process, the preset parsing model can progressively complete the processes of information extraction, rule retrieval, statement rewriting, data querying, and result interpretation (as well as code value querying) according to preset standardized processing steps. Each step corresponds to a tool call or a model inference in the toolchain, and there is a strict sequential dependency between the steps to fully reproduce the manual processing of verification issues. The toolchain can be a set of standardized business tools integrated into the preset parsing model, which may include tools for capturing the three elements of verification issues, rule knowledge base retrieval tools, SQL rewriting and data generation tools, code value table query tools, and free SQL execution tools, etc. Each tool encapsulates corresponding business logic for the model to call as needed.

[0094] Understandably, in this embodiment, the preset parsing model can be based on the ReAct (Reasoning and Acting) framework, following a preset six-step process to progressively think, invoke tools, and obtain results, thus advancing the parsing task step by step and achieving distributed parsing. Under this ReAct reasoning framework, the preset parsing model can first think to clarify the current goal, then decide to invoke the corresponding tools in the toolchain to obtain external information, and then proceed to the next step of thinking based on the results returned by the tools, iterating in a loop until the task is completed.

[0095] At this point, the initial analysis content mentioned above can be the original output set generated after the model completes all step-by-step reasoning and tool calls. It may contain intermediate results and preliminary conclusions of each step, but has not yet undergone compliance verification and adjustment.

[0096] Step S12: Adjust the initial parsed content for compliance based on preset business constraints to obtain the verification parsing result.

[0097] It is easy to understand that the aforementioned preset business constraints can be a pre-configured set of mandatory rules, covering multiple dimensions of rule information such as data accuracy, SQL standardization, process correctness, business compliance, or output format, to ensure that the model output meets the core constraints of business requirements. Therefore, the aforementioned compliance adjustment can refer to the process of verifying the initial parsed content item by item based on the preset business constraints, correcting, supplementing, or eliminating content that does not meet the preset business constraints, and ensuring that the final output is accurate, standardized, and compliant.

[0098] For example, in order to help understand the technical concept or technical principle of the data verification result parsing method after combining this embodiment with the above embodiments one and two, refer to Figure 4 Provide an explanation. Figure 4 The following is a simplified schematic diagram of the data verification result parsing method of this application: This embodiment proposes a complete framework for intelligent parsing of regulatory reporting data verification results, aiming to address the core issues of existing technologies where verification error parsing is highly reliant on manual intervention, inefficient, and has a high barrier to entry. Through a standardized four-step process of "rule matching - SQL optimization - data query - intelligent parsing," it automates and intelligently parses the causes of verification errors. This allows business personnel to quickly obtain the causes and details of errors without needing to understand the underlying SQL and technical logic, significantly lowering the threshold for verification processing, improving problem-solving efficiency, and reducing the repetitive workload of IT operations personnel in answering questions.

[0099] like Figure 4As shown, the parsing device in this solution can adopt a three-layer architecture design, including a verification display platform, an intelligent parsing module, and a database query module. The verification display platform serves as the user interaction entry point, receiving verification parsing requests containing the three verification elements and displaying the verification parsing results returned by the intelligent parsing module. The intelligent parsing module can integrate the WORKFLOW business orchestration mode corresponding to embodiments one and two above, i.e. Figure 4 The workflow parsing mode and the AGENT intelligent interaction mode proposed in this embodiment (i.e., the above-mentioned intelligent parsing mode) can be responsible for the core parsing logic processing and tool scheduling; while the database query module provides structured data access capabilities, executes SQL queries and returns results.

[0100] At this point, when a user submits a verification and parsing request on the verification and display platform and selects the intelligent parsing mode, the parsing device can first load the complete set of configurations for the preset parsing model as shown below: 1. Inject persona and task prompts, and create persona and response logic: Set the preset parsing model as an IT support expert with 5 years of experience in regulatory reporting and verification. The task is to clearly explain the reasons for verification errors to business personnel based on real data. The response should be professional, easy to understand, and in accordance with business terminology. 2. Register the toolchain for easy use of subsequent tools: Register tools such as verification issue element capture, rule knowledge base retrieval, verification execution statement rewriting and error detail generation, and code value knowledge base query to the model, and configure the input parameters, output parameters and calling methods of each tool; 3. Loading process rules: Clearly enforce the six-step process, in the following order: extract three elements → search rules → construct SQL → execute SQL → code value query → explain the reason. Skipping steps and reversing the order are prohibited. 4. Configure preset business constraints: All conclusions must be based on real data returned by the tool, and fabrication is prohibited; SQL must have its schema prefix removed; answers to questions unrelated to validation are prohibited.

[0101] Next, the parsing device can input the user's natural language request, i.e., the verification parsing request, into a preset parsing model to start the step-by-step parsing process, while simultaneously enabling streaming output to synchronize the execution progress with the user in real time. During the step-by-step execution of inference and tool calls by the parsing model to generate the initial parsed content, the "verification question element capture tool" can be invoked first. The user's question text is passed in, and the tool returns three core parameters extracted from the user's question: rule number, data period, and data block ID, which are the aforementioned verification question elements.

[0102] Then, the rule knowledge base is retrieved, that is, the "rule knowledge base retrieval tool" is invoked, the verification rule identifier and data period are passed in to retrieve the rule knowledge base, the rule semantic information, verification execution statement and verification fields returned by the tool are obtained, and after confirming that the rule is valid, the next step is taken.

[0103] Next, an optimized query SQL (i.e., the aforementioned optimized query statement) is constructed. This is done by calling the "Validation Execution Statement Rewriting and Error Details Generation Tool," which inputs the validation execution statement, validation fields, and primary key configuration, and retrieves the rewritten optimized query statement returned by the tool. After detecting that the validation optimized query statement is syntactically correct and confirming that the document to be accessed corresponding to the optimized query statement falls within the accessible document range corresponding to the validation object identifier, the optimized query statement is executed. The returned exception data information is then retrieved from the database using the validation object identifier and data period. Figure 4 The error details shown are as follows. This error details data may include the total number of errors and detailed / statistical results. The parsing device can proceed to the next step after confirming the validity of this data.

[0104] Optionally, the model can determine and query the code value table. If code value verification is involved in the data, the "Code Value Knowledge Base Query Tool" can be invoked, passing in the identified code value verification code. At this point, the model can load a code value-specific prompt word template. This prompt word template can fill the corresponding positions in the model's output data template with rule semantic information, abnormal data information, and code value example data. The prompt word constraint model output can include the following: 1. Rule requirement description: clearly stating that the field must conform to the standard code value specification of the corresponding type; 2. Overall error situation: stating the total number of errors and the distribution statistics of each error value; 3. Error cause analysis: clearly indicating that the field value of the data is not within the standard code value range and is an invalid code; 4. Compliant code value example: listing the correct code values ​​found and their corresponding names for user reference; 5. Rectification direction suggestions: prompting the user to correct the encoding of the corresponding field according to the standard code value table. If there is no code value verification, skip this step and proceed to the next step.

[0105] Next, the model can generate an explanation of the verification error based on the prompt word template. First, based on the rule requirements and actual abnormal data, the model analyzes the specific reasons why the data triggers the verification, organizes the language according to structured requirements, including rule descriptions, error situations, detailed examples, and a summary of reasons, and completes the final conclusion generation.

[0106] Finally, after the preset parsing model completes all the above steps, the execution records of each step, the tool's returned results, and the final interpretation can be summarized and assembled into complete initial parsing content. Figure 4The intelligent response summary process shown loads all preset business constraints and automatically verifies and adjusts the initial parsed content item by item to obtain the final verification and parsing result. Data verification items that can be performed during this process include: 1) Data accuracy verification: comparing the data descriptions in the explanation content with the actual abnormal data returned by the tool, correcting statements that do not match the actual data, and deleting subjective conclusions without data support; 2) SQL standardization verification: checking the output SQL statements, automatically removing the schema prefix before table names to ensure standardized SQL format; 3) Process correctness verification: verifying the execution order and completeness of steps, and automatically scheduling the model to supplement the execution of corresponding steps if skipped steps exist; 4) Business compliance verification: checking whether the output content contains illegal statements or irrelevant content, correcting non-standard business terminology, and ensuring that the output focuses on the verification issue itself; 5) Format standardization verification: adjusting the output structure and format to ensure compliance with front-end display requirements.

[0107] In addition, such as Figure 4 As shown, this embodiment also supports multi-round interaction and extended question answering. After viewing the parsing results, users can continue to ask follow-up questions through the verification display platform, such as "What is the definition of this field?", "How do I modify this data?", and "What other similar verifications are there?". At this time, the preset parsing model can also determine the question type and execute corresponding processing steps based on the contextual dialogue information. If the question is answerable in the knowledge base, the corresponding knowledge base query tool can be directly invoked to obtain the information and reply; if it is a data-related question, the corresponding verification execution statement rewriting and error detail generation tool in the toolchain can be invoked to query the database to obtain the data and reply; if it exceeds the system's capabilities, the user can be clearly informed and advised to contact the relevant responsible personnel. Furthermore, multi-round dialogues retain the entire context, supporting continuous follow-up questions and achieving an immersive Q&A experience.

[0108] Furthermore, this embodiment can also set up a multi-layered exception fallback mechanism to ensure process stability. Specifically, if a tool call fails, such as two consecutive failed calls to a single tool, the model can explain the current stage to the user, suggesting switching to workflow parsing mode or retrying later. If a model inference timeout occurs, such as a single-step inference exceeding a preset 30-second timeout threshold, the system can automatically terminate the parsing, return a timeout message, and provide a retry entry. If parameters are missing, such as incomplete extraction of the three verification elements, the model can proactively ask the user for the missing information and continue executing the complete process after the user provides the missing details.

[0109] Therefore, to address the problem that existing fixed workflow parsing models only output the final result, making the analysis process black-box, business personnel only know what happens but not why, unable to understand the verification logic and problem localization methods, and thus unable to accumulate and pass on knowledge, resulting in extremely high overall learning costs, this embodiment can fully reproduce the processing ideas and operation steps of IT experts through intelligent parsing mode. Business personnel can learn the verification logic and problem localization methods independently by viewing the process, improving the efficiency of experience transfer and effectively solving the knowledge gap problem caused by personnel turnover.

[0110] Furthermore, this embodiment upgrades the data parsing process from "providing answers" to "teaching methods," helping business personnel gradually master the ability to handle verification issues, shifting from passively receiving results to proactively solving problems. In the long run, this can significantly reduce reliance on IT personnel and empower business personnel with autonomy. Therefore, the intelligent parsing mode proposed in this embodiment can perfectly complement the workflow parsing solidification mode, comprehensively covering the needs of different business scenarios.

[0111] This embodiment discloses that, when the current parsing mode is intelligent parsing mode, the user-inputted verification parsing request is input into a preset parsing model for step-by-step parsing to obtain initial parsing content. The initial parsing content includes at least one of the following: verification problem elements, verification rule information, optimized query statements, code value example data, and abnormal data information. Based on preset business constraints, the initial parsing content is adjusted for compliance to obtain the verification parsing result.

[0112] This embodiment simulates the thinking and operation process of IT experts, making the parsing process fully transparent and explainable, helping business personnel understand the verification logic and master the processing methods, and effectively realizing knowledge transfer. At the same time, this embodiment supports multi-round interaction and extended question answering, empowering business personnel to grow their capabilities while further reducing the Q&A burden on IT personnel and covering a more comprehensive range of business scenarios.

[0113] It should be noted that the above examples are only for understanding this application and do not constitute a limitation on the data verification result parsing method of this application. Any simple modifications based on this technical concept are within the protection scope of this application.

[0114] This application also provides a data verification result parsing device, please refer to... Figure 5 , Figure 5 This is a schematic diagram of the module structure of the data verification result parsing device according to an embodiment of this application. In this embodiment, the data verification result parsing device includes: The rule query module 501 is used to perform rule matching on the user-input verification parsing request to obtain the verification object identifier and verification rule information, wherein the verification rule information includes rule semantic information and verification execution statement; Statement optimization module 502 is used to optimize the fields of the verification execution statement according to the rule semantic information to obtain an optimized query statement; Anomaly detection module 503 is used to obtain abnormal data information based on the optimized query statement and the verification object identifier; The data parsing module 504 is used to generate a verification parsing result based on the rule semantic information and the abnormal data information.

[0115] The data verification result parsing device provided in this application, employing the data verification result parsing method in the above embodiments, can solve the technical problem of how to improve the efficiency of verification anomaly handling. Compared with the prior art, the beneficial effects of the data verification result parsing device provided in this application are the same as those of the data verification result parsing method provided in the above embodiments, and other technical features in the data verification result parsing device are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0116] This application provides a data verification result parsing device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the data verification result parsing method in the first embodiment described above.

[0117] The following is for reference. Figure 6 The diagram illustrates a structural schematic of a data verification result parsing device suitable for implementing embodiments of this application. The data verification result parsing device in embodiments of this application may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Description), PMPs (Portable Media Players), in-vehicle terminals (e.g., in-vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. Figure 6 The data verification result parsing device shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this application.

[0118] like Figure 6As shown, the data verification result parsing device may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1002 or a program loaded from a storage device 1003 into a random access memory (RAM) 1004. The RAM 1004 also stores various programs and data required for the operation of the data verification result parsing device. The processing unit 1001, the ROM 1002, and the RAM 1004 are interconnected via a bus 1005. An input / output interface 1006 is also connected to the bus. Typically, the following systems can be connected to the input / output interface 1006: input devices 1007 including, for example, touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices 1008 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 1003 including, for example, magnetic tapes, hard disks, etc.; and communication devices 1009. Communication device 1009 allows the data verification result parsing device to communicate wirelessly or wiredly with other devices to exchange data. Although the figure shows data verification result parsing devices with various systems, it should be understood that implementation or possession of all the systems shown is not required. More or fewer systems may be implemented alternatively.

[0119] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, an embodiment disclosed in this application includes a data verification result parsing program product, which includes a data verification result parsing program carried on a computer-readable medium, the data verification result parsing program containing program code for performing the methods shown in the flowcharts. In such an embodiment, the data verification result parsing program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from read-only memory 1002. When the data verification result parsing program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.

[0120] The data verification result parsing device provided in this application, employing the data verification result parsing method described in the above embodiments, can solve the technical problem of how to improve the efficiency of verification anomaly handling. Compared with the prior art, the beneficial effects of the data verification result parsing device provided in this application are the same as those of the data verification result parsing method described in the above embodiments, and other technical features in this data verification result parsing device are the same as those disclosed in the method of the previous embodiment, and will not be repeated here.

[0121] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0122] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0123] This application provides a storage medium having computer-readable program instructions (i.e., a data verification result parsing program) stored thereon, which are used to execute the data verification result parsing method in the above embodiments.

[0124] The storage medium provided in this application may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or any combination thereof. More specific examples of the storage medium may include, but are not limited to: electrical connections with one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, system, or device. The program code contained on the storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0125] The aforementioned storage medium may be included in the data verification result parsing device; or it may exist independently and not be assembled into the data verification result parsing device.

[0126] The aforementioned storage medium carries one or more programs. When the aforementioned one or more programs are executed by the data verification result parsing device, the data verification result parsing device is enabled to perform data verification result parsing.

[0127] Data verification result parsing program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof. These programming languages ​​include object-oriented programming languages—such as Java, Smalltalk, and C++—as well as conventional procedural programming languages—such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a Local Area Network (LAN) or a Wide Area Network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0128] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and data verification result parsing 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 the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than those shown 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 the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can 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.

[0129] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0130] The readable storage medium provided in this application is a storage medium that stores computer-readable program instructions (i.e., a data verification result parsing program) for executing the above-described data verification result parsing method, thereby solving the technical problem of how to improve the efficiency of verification anomaly handling. Compared with the prior art, the beneficial effects of the storage medium provided in this application are the same as the beneficial effects of the data verification result parsing method provided in the above embodiments, and will not be repeated here.

[0131] The above are only some embodiments of this application and do not limit the scope of the solution of this application. All equivalent structural transformations made under the technical concept of this application and using the content of this application specification and drawings, or direct / indirect applications in other related technical fields, are included within the protection scope of this application.

Claims

1. A method for parsing data verification results, characterized in that, The method includes: The user-input verification parsing request is matched with rules to obtain the verification object identifier and verification rule information, wherein the verification rule information includes rule semantic information and verification execution statements; Based on the semantic information of the rules, the validation execution statement is optimized to obtain an optimized query statement; Abnormal data information is obtained based on the optimized query statement and the verification object identifier; A verification and parsing result is generated based on the semantic information of the rules and the abnormal data information.

2. The data verification result parsing method as described in claim 1, characterized in that, The step of optimizing the fields of the validation execution statement based on the rule semantic information to obtain an optimized query statement includes: The key fields of the rule semantic information and the verification execution statement are parsed to obtain the verification-related field information; The execution subject of the verification execution statement is parsed to obtain the query subject information; Based on the query subject information and the verification related field information, the range of returned fields of the verification execution statement is adjusted to obtain an optimized query statement.

3. The data verification result parsing method as described in claim 1, characterized in that, The step of performing rule matching on the user-input verification parsing request to obtain the verification object identifier and verification rule information includes: Semantic parsing is performed on the user-inputted verification parsing request to obtain verification question elements, which include: data period information, verification object identifier, and verification rule identifier; Based on the data period information and the verification rule identifier, the verification rule knowledge base is queried to obtain the verification rule information.

4. The data verification result parsing method as described in claim 1, characterized in that, The step of obtaining abnormal data information based on the optimized query statement and the verification object identifier includes: Based on the verification object identifier, the optimized query statement is executed with access restrictions to obtain the initial query result; Obtain the data volume characteristics of the initial query results, and determine the current output strategy based on the data volume characteristics; The initial query results are processed based on the current output strategy to obtain abnormal data information.

5. The data verification result parsing method as described in claim 4, characterized in that, The step of restricting access permissions for the optimized query statement based on the verification object identifier to obtain the initial query result includes: Determine whether the document to be accessed corresponding to the optimized query statement belongs to the accessible document range corresponding to the verification object identifier; If it is detected that the document to be accessed belongs to the range of accessible documents, the output restriction execution result corresponding to the optimized query statement is used as the initial query result.

6. The data verification result parsing method as described in claim 1, characterized in that, The step of generating a verification and parsing result based on the rule semantic information and the abnormal data information includes: Detect whether there is a code value verification code in the optimized query statement; If so, then query the code value configuration table based on the code value verification encoding to obtain code value example data; A verification and parsing result is generated based on the code value example data, the rule semantic information, and the abnormal data information.

7. The data verification result parsing method as described in claim 1, characterized in that, The method further includes: When the current parsing mode is intelligent parsing mode, the user-inputted verification parsing request is input into a preset parsing model for step-by-step parsing to obtain initial parsing content. The initial parsing content includes at least one of the following: verification problem elements, verification rule information, optimized query statements, code value example data, and abnormal data information. The initial parsed content is adjusted for compliance based on preset business constraints to obtain the verification parsing result.

8. A data verification result parsing device, characterized in that, The data verification result parsing device includes: The rule query module is used to perform rule matching on the user-input verification parsing request to obtain the verification object identifier and verification rule information, wherein the verification rule information includes rule semantic information and verification execution statement; The statement optimization module is used to optimize the fields of the validation execution statement based on the rule semantic information to obtain an optimized query statement; An anomaly detection module is used to obtain abnormal data information based on the optimized query statement and the verification object identifier; The data parsing module is used to generate verification and parsing results based on the rule semantic information and the abnormal data information.

9. A data verification result parsing device, characterized in that, The device includes: a memory, a processor, and a data verification result parsing program stored in the memory and executable on the processor, the data verification result parsing program being configured to implement the steps of the data verification result parsing method as described in any one of claims 1 to 7.

10. A storage medium, characterized in that, The storage medium stores a data verification result parsing program, which, when executed by a processor, implements the steps of the data verification result parsing method as described in any one of claims 1 to 7.

Images