Method, system and equipment for automatically generating rectification measure based on enterprise inspection problem, and medium

Abstract

The invention discloses a rectification measure automatic generation method, a rectification measure automatic generation system, rectification measure automatic generation equipment and a medium based on an enterprise patrol problem, belongs to the technical field of data processing and natural language processing, and aims to solve the technical problem of how to realize rapid analysis of the enterprise patrol problem and rectification measure automatic generation. According to the technical scheme, the method comprises the following steps: problem classification and feature extraction: analyzing a lexical method, a syntax method and a semantic structure of a patrol problem text through an NLP model, identifying a problem type, determining a related field, evaluating a problem severity degree, and extracting keywords; meanwhile, a set of complete inspection question label system is constructed, multi-dimensional labeling is conducted on the question, and the form is as follows: label dimensions comprise question types, related fields, question severity and keywords; constructing a knowledge graph; creating a rule base and generating candidate rectification measures; and generating and optimizing rectification measures.

Description

Technical Field

[0001] This invention relates to the fields of data processing and natural language processing technology, specifically to a method, system, device, and medium for automatically generating rectification measures based on enterprise inspection issues. Background Technology

[0002] Traditional methods of handling such issues mainly rely on manual operation. Inspection staff need to analyze each inspection problem and formulate rectification measures based on their own experience, relevant regulations and policies, and historical cases.

[0003] In recent years, although some studies have attempted to use technology to assist inspection work, existing technologies are mostly limited to simple data storage and retrieval, or use a single algorithm to perform preliminary classification of problems. They are unable to systematically generate comprehensive, accurate and compliant rectification measures based on the characteristics of the inspection problems.

[0004] Therefore, how to achieve rapid analysis of problems found during enterprise inspections and the automated generation of rectification measures, thereby improving the scientific nature and standardization of inspection and rectification work, is a technical problem that urgently needs to be solved. Summary of the Invention

[0005] The technical objective of this invention is to provide a method, system, equipment, and medium for automatically generating rectification measures based on enterprise inspection issues, in order to address the problem of how to achieve rapid analysis of enterprise inspection issues and automated generation of rectification measures, thereby improving the scientific and standardized nature of inspection and rectification work.

[0006] The technical objective of this invention is achieved as follows: a method for automatically generating corrective measures based on issues identified during enterprise inspections, the method being as follows:

[0007] Problem Classification and Feature Extraction: The lexical, syntactic, and semantic structures of inspection problem texts are analyzed using an NLP model to identify problem types, determine the relevant domains, assess problem severity, and extract keywords. Simultaneously, a complete inspection problem tagging system is constructed to label problems from multiple dimensions, in the form of: tag dimensions including problem type + relevant domain + problem severity + keywords; an example is "Organizational Management Problem - Human Resources Management - Minor - Talent Loss".

[0008] Building a knowledge graph: Collecting and integrating company rules and regulations as well as historical corporate inspection cases to build a knowledge graph, enabling structured, interconnected storage and efficient retrieval of various types of information;

[0009] Create a rule base and generate candidate rectification measures: Create a rule base based on a knowledge graph, and generate candidate rectification measures through "problem tag matching + rule priority sorting + case similarity calculation";

[0010] Generate and optimize rectification measures: Candidate rectification measures generated by matching with the rule base are generated using a template-based method. Standardized rectification measures are output based on standardized language templates. The generated rectification measures are then modified in terms of content, priority, and details such as adding responsible persons and time limits. At the same time, reinforcement learning algorithms are used to automatically optimize rule weights based on feedback from historical rectification results. For example, if a certain type of measure has a high success rate in rectification, its recommendation priority is increased.

[0011] As a preferred approach, the problem types include organizational management issues, operational management issues, and financial management issues; the areas include finance, human resources, and general management; the severity of the problems includes minor, moderate, and serious; and the keywords include irregular expense reimbursement, personnel selection and appointment, and lack of supervision.

[0012] As a preferred approach, the nodes in the knowledge graph cover key information such as problem type, problem cause, rectification measures, responsible department, and rules and regulations. The relationships between nodes in the knowledge graph include problem → cause (e.g., "illegal procurement" → "lack of internal control mechanism"), cause → measure (e.g., "lack of internal control mechanism" → "improving procurement management system"), and measure → regulations (e.g., "improving procurement management system" → "Enterprise Procurement Standard").

[0013] As a preferred option, the rule base includes severity rules, domain rules, and repetitive rules;

[0014] Among them, the severity rule stipulates that if the problem label includes "supply chain disruption risk", then the corrective measures must include "supplier diversification management" and "digital supply chain monitoring".

[0015] The rules stipulate that if the issue involves "finance - fund management", the corrective measures will be automatically linked to "financial audit", "account verification" and "improving reimbursement system".

[0016] The repetitive rule stipulates that if similar problems recur in historical cases, the rectification measures will include "establishing a long-term supervision mechanism" and "interviewing the responsible persons".

[0017] As a preferred option, the candidate rectification measures are as follows:

[0018] Get the question and extract its tags;

[0019] Matching and retrieval are performed in the rule base and knowledge graph;

[0020] Candidate rectification measures were selected based on rule priority and case similarity.

[0021] More preferably, the candidate rectification measures generated through rule base matching are generated using a template-based approach, and the standardized rectification measures are output based on standardized language templates as follows:

[0022] The rectification measures adopt the template of "In response to the XX problem, the XX department needs to complete the XX rectification (such as revising the system, optimizing the channel, etc.) within the XX time limit".

[0023] Accountability measures use the template "The person responsible for the XX issue will be given XX disciplinary action (such as warning, performance deduction, etc.) by the XX department in accordance with regulations and rules."

[0024] Preventive measures should follow the template of "To prevent similar problems from happening again, it is necessary to establish XX mechanism (such as regular audits, normalized legal education, etc.)".

[0025] An automatic generation system for corrective measures based on issues identified during enterprise inspections, comprising:

[0026] The problem classification and feature extraction module is used to analyze the lexical, syntactic, and semantic structure of inspection problem texts using an NLP model, identify problem types, determine the relevant domains, assess the severity of problems, and extract keywords. It also constructs a complete inspection problem tagging system to label problems from multiple dimensions, in the form of: tag dimensions including problem type + relevant domain + problem severity + keywords; an example is "Organizational Management Problem - Human Resources Management - Minor - Talent Loss".

[0027] The knowledge graph construction module is used to collect and integrate company rules and regulations as well as historical corporate inspection cases to build a knowledge graph, enabling structured and interconnected storage and efficient retrieval of various types of information;

[0028] The knowledge graph creation and measure generation module is used to create a rule base based on the knowledge graph and generate candidate rectification measures through "problem tag matching + rule priority sorting + case similarity calculation".

[0029] The rectification measures generation and optimization module is used to match candidate rectification measures generated through a rule base. It adopts a template-based generation method, outputs standardized rectification measures based on standardized language templates, and makes detailed optimizations to the generated rectification measures, such as modifying the content, adjusting the priority, and adding responsible persons and time limits. At the same time, it combines reinforcement learning algorithms to automatically optimize rule weights based on historical rectification effect feedback. For example, if a certain type of measure has a high rectification success rate, its recommendation priority will be increased.

[0030] As a preferred approach, the problem types include organizational management issues, operational management issues, and financial management issues; the areas include finance, human resources, and general management; the severity of the problems includes minor, moderate, and serious; and the keywords include irregular expense reimbursement, personnel selection and appointment, and lack of supervision.

[0031] The nodes in the knowledge graph cover key information such as problem type, problem cause, rectification measures, responsible department, and rules and regulations; the relationships between nodes in the knowledge graph include problem → cause (e.g., "illegal procurement" → "lack of internal control mechanism"), cause → measure (e.g., "lack of internal control mechanism" → "improving procurement management system"), and measure → regulations (e.g., "improving procurement management system" → "Enterprise Procurement Standards").

[0032] The rule base includes severity rules, domain rules, and recurrence rules. Among them, the severity rules stipulate that if the problem label includes "supply chain disruption risk," the corrective measures must include "supplier diversification management" and "digital supply chain monitoring." The domain rules stipulate that if the problem involves "financial domain - fund management," the corrective measures will automatically be associated with "financial audit," "account verification," and "improving the reimbursement system." The recurrence rules stipulate that if similar problems recur in historical cases, the corrective measures will include "establishing a long-term supervision mechanism" and "interviewing the responsible person."

[0033] The process of generating candidate rectification measures is as follows: first, the problem is identified and its tags are extracted; then, matching and retrieval are performed in the rule base and knowledge graph; finally, corresponding candidate rectification measures are selected according to rule priority and case similarity.

[0034] The specific rectification measures based on the standardized language template output specifications are as follows: Rectification measures adopt the template "Regarding the XX problem, the XX department needs to complete the XX rectification (such as revising the system, optimizing the channel, etc.) within the XX time limit"; Accountability measures use the template "The XX department shall give XX punishment (such as warning, deduction of performance, etc.) to the relevant responsible persons for the XX problem in accordance with discipline and regulations"; Prevention measures adopt the template "In order to avoid similar problems from happening again, it is necessary to establish the XX mechanism (such as regular audits, normalization of legal education, etc.)".

[0035] An electronic device includes: a memory and at least one processor;

[0036] The memory contains computer programs;

[0037] The at least one processor executes the computer program stored in the memory, causing the at least one processor to perform the automatic generation method for corrective measures based on enterprise inspection issues as described above.

[0038] A computer-readable storage medium storing a computer program that can be executed by a processor to implement the above-described method for automatically generating corrective measures based on enterprise inspection issues.

[0039] The automatic generation method, system, equipment, and medium for corrective measures based on enterprise inspection issues of the present invention have the following advantages:

[0040] (i) This invention integrates technologies such as natural language processing, knowledge graphs and rule engines to achieve rapid analysis of inspection issues and automated generation of rectification measures, solving the problems of low efficiency and inconsistent standards in manual processing, and improving the scientific and standardized nature of inspection rectification work;

[0041] (ii) Compared with the traditional method of manually formulating rectification measures, the present invention can quickly analyze the inspection problems and generate rectification measures, which greatly shortens the time for formulating rectification plans and effectively improves the response speed of inspection work;

[0042] (III) This invention, through automated processing based on knowledge graphs and rule engines, ensures that rectification measures for similar problems follow unified standards and norms, avoids inconsistencies in processing due to the subjectivity of manual operation, improves the standardization and scientific nature of inspection and rectification work, and ensures unified rectification standards for enterprises.

[0043] (iv) The measures of this invention are all based on regulations and historical experience, which can effectively avoid compliance risks caused by unreasonable measures, ensure that the inspection and rectification work is carried out within a legal and compliant framework, and effectively reduce the risk of enterprise rectification.

[0044] (V) This invention is deployed in the inspection and rectification system for integrated use by the comprehensive management and supervision system. System users can input inspection problems by text or voice, and click the intelligent recognition button. The system will search the knowledge graph and rule engine to match the 10 best rectification measures (including rectification measures, recommendation degree, etc., and arranged in descending order of recommendation degree). After selecting a rectification measure, compliance verification and conflict detection will be called. If it passes, a structured report will be generated. If it fails, the rectification measures can be modified or reselected according to the prompt information.

[0045] (vi) This invention uses knowledge graph retrieval and rule base to automatically match rectification measures, and uses compliance verification and conflict detection to verify the rationality of rectification measures, which is used in scenarios where enterprises can plan and generate problem rectification measures in a timely manner. Attached Figure Description

[0046] The invention will be further described below with reference to the accompanying drawings.

[0047] Appendix Figure 1 A flowchart illustrating the method for automatically generating corrective measures based on issues identified during enterprise inspections. Detailed Implementation

[0048] The following detailed description of the automatic generation method, system, equipment, and medium for corrective measures based on enterprise inspection issues, with reference to the accompanying drawings and specific embodiments, is provided in the specification.

[0049] Example 1:

[0050] As attached Figure 1 As shown in the figure, this embodiment provides a method for automatically generating rectification measures based on enterprise inspection issues. The method is as follows:

[0051] S1. Problem Classification and Feature Extraction: The lexical, syntactic, and semantic structures of inspection problem texts are analyzed using an NLP model to identify problem types, determine the relevant domains, assess problem severity, and extract keywords. Simultaneously, a complete inspection problem tagging system is constructed to label problems from multiple dimensions, in the form of: tag dimensions including problem type + relevant domain + problem severity + keywords; an example is "Organizational Management Problem - Human Resources Management - Minor - Talent Loss".

[0052] S2. Construct a knowledge graph: Collect and integrate company rules and regulations and historical corporate inspection cases to construct a knowledge graph, enabling structured and interconnected storage and efficient retrieval of various types of information;

[0053] S3. Create a rule base and generate candidate rectification measures: Create a rule base based on the knowledge graph, and generate candidate rectification measures through "problem tag matching + rule priority sorting + case similarity calculation";

[0054] S4. Generate and optimize rectification measures: Candidate rectification measures generated by matching with the rule base are generated using a template-based method. Standardized rectification measures are output based on standardized language templates. The generated rectification measures are then modified in terms of content, priority, and details such as adding responsible persons and time limits. At the same time, reinforcement learning algorithms are used to automatically optimize rule weights based on feedback from historical rectification results. For example, if a certain type of measure has a high success rate in rectification, its recommendation priority is increased.

[0055] The problem types in step S1 of this embodiment include organizational management problems, operational management problems, and financial management problems; the areas include finance, human resources, and general management; the severity of the problems includes minor, general, and serious; and the keywords include irregular expense reimbursement, personnel selection and appointment, and lack of supervision.

[0056] In step S2 of this embodiment, the nodes in the knowledge graph cover key information such as problem type, problem cause, rectification measures, responsible department, and rules and regulations. The relationships between the nodes in the knowledge graph include problem → cause (e.g., "illegal procurement" → "lack of internal control mechanism"), cause → measure (e.g., "lack of internal control mechanism" → "improving procurement management system"), and measure → regulations (e.g., "improving procurement management system" → "Enterprise Procurement Standard").

[0057] The rule base in step S3 of this embodiment includes severity rules, domain rules, and repetitive rules;

[0058] Among them, the severity rule stipulates that if the problem label includes "supply chain disruption risk", then the corrective measures must include "supplier diversification management" and "digital supply chain monitoring".

[0059] The rules stipulate that if the issue involves "finance - fund management", the corrective measures will be automatically linked to "financial audit", "account verification" and "improving reimbursement system".

[0060] The repetitive rule stipulates that if similar problems recur in historical cases, the rectification measures will include "establishing a long-term supervision mechanism" and "interviewing the responsible persons".

[0061] The specific steps for generating candidate rectification measures in step S3 of this embodiment are as follows:

[0062] S301. Obtain the issue and extract the issue tags;

[0063] S302. Perform matching and retrieval in the rule base and knowledge graph;

[0064] S303. Select corresponding candidate rectification measures according to rule priority and case similarity.

[0065] In step S4 of this embodiment, the candidate rectification measures generated by matching the rule base are generated using a template-based method. The specific rectification measures output based on standardized language templates are as follows:

[0066] ① Rectification measures adopt the template "In response to the XX problem, the XX department needs to complete the XX rectification (such as revising the system, optimizing the channel, etc.) within the XX time limit";

[0067] ② For accountability measures, use the template "The person responsible for the XX issue will be given XX punishment (such as warning, performance deduction, etc.) by the XX department in accordance with discipline and regulations."

[0068] ③ For preventative measures, use the template "To prevent similar problems from happening again, it is necessary to establish XX mechanism (such as regular audits, normalized legal education, etc.)".

[0069] Example 2:

[0070] This embodiment provides an automatic generation system for corrective measures based on enterprise inspection issues. The system includes:

[0071] The problem classification and feature extraction module is used to analyze the lexical, syntactic, and semantic structure of inspection problem texts using an NLP model, identify problem types, determine the relevant domains, assess the severity of problems, and extract keywords. It also constructs a complete inspection problem tagging system to label problems from multiple dimensions, in the form of: tag dimensions including problem type + relevant domain + problem severity + keywords; an example is "Organizational Management Problem - Human Resources Management - Minor - Talent Loss".

[0072] The knowledge graph construction module is used to collect and integrate company rules and regulations as well as historical corporate inspection cases to build a knowledge graph, enabling structured and interconnected storage and efficient retrieval of various types of information;

[0073] The knowledge graph creation and measure generation module is used to create a rule base based on the knowledge graph and generate candidate rectification measures through "problem tag matching + rule priority sorting + case similarity calculation".

[0074] The rectification measures generation and optimization module is used to match candidate rectification measures generated through a rule base. It adopts a template-based generation method, outputs standardized rectification measures based on standardized language templates, and makes detailed optimizations to the generated rectification measures, such as modifying the content, adjusting the priority, and adding responsible persons and time limits. At the same time, it combines reinforcement learning algorithms to automatically optimize rule weights based on historical rectification effect feedback. For example, if a certain type of measure has a high rectification success rate, its recommendation priority will be increased.

[0075] In this embodiment, the problem types include organizational management problems, operational management problems, and financial management problems; the areas include finance, human resources, and general management; the severity of the problems includes minor, moderate, and severe; and the keywords include irregular expense reimbursement, personnel selection and appointment, and lack of supervision.

[0076] In this embodiment, the nodes in the knowledge graph cover key information such as problem type, problem cause, rectification measures, responsible department, and rules and regulations. The relationships between the nodes in the knowledge graph include problem → cause (e.g., "illegal procurement" → "lack of internal control mechanism"), cause → measure (e.g., "lack of internal control mechanism" → "improving procurement management system"), and measure → regulations (e.g., "improving procurement management system" → "Enterprise Procurement Standard").

[0077] In this embodiment, the rule base includes severity rules, domain rules, and repetitive rules. Among them, the severity rules stipulate that if the problem label includes "supply chain disruption risk", the rectification measures must include "supplier diversification management" and "digital supply chain monitoring". The domain rules stipulate that if the problem involves "financial domain - fund management", the rectification measures will automatically be associated with "financial audit", "account verification" and "improving reimbursement system". The repetitive rules stipulate that if similar problems recur in historical cases, the rectification measures will be superimposed with "establishing a long-term supervision mechanism" and "interviewing the responsible person".

[0078] In this embodiment, generating candidate rectification measures specifically involves: first, obtaining the problem and extracting the problem tag; then, matching and searching in the rule base and knowledge graph; and finally, selecting the corresponding candidate rectification measures according to rule priority and case similarity.

[0079] In this embodiment, the rectification measures based on the standardized language template output specifications are as follows: Rectification measures adopt the template "For the XX problem, the XX department needs to complete the XX rectification (such as revising the system, optimizing the channel, etc.) within the XX time limit"; accountability measures use the template "For the relevant responsible persons of the XX problem, the XX department shall give XX punishment (such as warning, deduction of performance, etc.) in accordance with discipline and regulations"; and prevention measures adopt the template "In order to avoid similar problems from happening again, it is necessary to establish the XX mechanism (such as regular auditing, normalization of legal education, etc.)".

[0080] Example 3:

[0081] This embodiment also provides an electronic device, including: a memory and a processor;

[0082] The memory stores the instructions executed by the computer.

[0083] The processor executes computer execution instructions stored in the memory, causing the processor to execute the automatic generation method for rectification measures based on enterprise inspection problems in any embodiment of the present invention.

[0084] The processor can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), off-the-shelf programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The processor can be a microprocessor or any conventional processor.

[0085] Memory is used to store computer programs and / or modules. The processor implements various functions of the electronic device by running or executing the computer programs and / or modules stored in the memory, and by accessing data stored in the memory. Memory can mainly include a program storage area and a data storage area. The program storage area can store the operating system, at least one application program required for a function, etc.; the data storage area can store data created based on the use of the terminal, etc. In addition, memory can also include high-speed random access memory, and can also include non-volatile memory, such as hard disks, RAM, plug-in hard disks, smart memory cards (SMC), secure digital cards (SD cards), flash memory cards, at least one disk storage device, flash memory devices, or other volatile solid-state storage devices.

[0086] Example 4:

[0087] This embodiment also provides a computer-readable storage medium storing multiple instructions, which are loaded by a processor to cause the processor to execute the automatic generation method for rectification measures based on enterprise inspection problems in any embodiment of the present invention. Specifically, a system or apparatus equipped with a storage medium may be provided, on which software program code implementing the functions of any of the above embodiments is stored, and the computer (or CPU or MPU) of the system or apparatus can read and execute the program code stored in the storage medium.

[0088] In this case, the program code read from the storage medium can itself implement the function of any of the above embodiments, and therefore the program code and the storage medium storing the program code constitute part of the present invention.

[0089] Storage media embodiments for providing program code include floppy disks, hard disks, magneto-optical disks, optical disks (such as CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RYM, DVD-RW, DVD+RW), magnetic tapes, non-volatile memory cards, and ROMs. Alternatively, program code can be downloaded from a server computer via a communication network.

[0090] Furthermore, it should be clear that not only can the program code read by the computer be executed, but also the operating system or other components operating on the computer can be instructed based on the program code to perform some or all of the actual operations, thereby realizing the function of any of the embodiments described above.

[0091] Furthermore, it is understood that the program code read from the storage medium is written to the memory set in the expansion board inserted into the computer or to the memory set in the expansion unit connected to the computer. Then, based on the instructions of the program code, the CPU or other components installed on the expansion board or expansion unit execute some and all of the actual operations, thereby realizing the function of any of the embodiments described above.

[0092] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for automatically generating rectification measures based on enterprise inspection issues, characterized in that, The method is as follows: Problem classification and feature extraction: The lexical, syntactic and semantic structure of the inspection problem text is analyzed using an NLP model to identify problem types, determine the domains involved, assess the severity of the problems, and extract keywords; at the same time, a complete inspection problem labeling system is constructed to label problems in multiple dimensions, in the form of: label dimensions including problem type + domain involved + problem severity + keywords; Building a knowledge graph: Collecting and integrating company rules and regulations as well as historical corporate inspection cases to build a knowledge graph, enabling structured, interconnected storage and efficient retrieval of various types of information; Create a rule base and generate candidate rectification measures: Create a rule base based on knowledge graph, and generate candidate rectification measures through "problem tag matching + rule priority sorting + case similarity calculation"; Generate and optimize rectification measures: Candidate rectification measures generated by matching with the rule base are generated using a template-based method. Standardized rectification measures are output based on standardized language templates. The generated rectification measures are then modified in terms of content, priority, and details such as adding responsible persons and time limits. At the same time, reinforcement learning algorithms are used to automatically optimize rule weights based on feedback from historical rectification effects.

2. The method for automatically generating rectification measures based on enterprise inspection issues according to claim 1, characterized in that, Problem types include organizational management issues, operational management issues, and financial management issues; areas include finance, human resources, and general management; problem severity includes minor, moderate, and severe; keywords include irregular expense reimbursement, personnel selection and appointment, and lack of supervision.

3. The method for automatically generating rectification measures based on enterprise inspection issues according to claim 1, characterized in that, The nodes in the knowledge graph cover key information such as problem type, problem cause, rectification measures, responsible department, and rules and regulations; the relationships between nodes in the knowledge graph include problem → cause, cause → measure, and measure → rule.

4. The method for automatically generating rectification measures based on enterprise inspection issues according to claim 1, characterized in that, The rule base includes severity rules, domain rules, and repetitive rules; Among them, the severity rule stipulates that if the problem label includes "supply chain disruption risk", then the corrective measures must include "supplier diversification management" and "digital supply chain monitoring". The rules stipulate that if the issue involves "financial domain - fund management", the corrective measures will be automatically linked to "financial audit", "account verification" and "improving reimbursement system"; The repetitive rule stipulates that if similar problems recur in historical cases, the rectification measures will include "establishing a long-term supervision mechanism" and "interviewing the responsible persons".

5. The method for automatically generating rectification measures based on enterprise inspection issues according to claim 1, characterized in that, The specific candidate rectification measures generated are as follows: Get the question and extract its tags; Matching and retrieval are performed in the rule base and knowledge graph; Candidate rectification measures were selected based on rule priority and case similarity.

6. The method for automatically generating rectification measures based on enterprise inspection issues according to any one of claims 1-5, characterized in that, The candidate rectification measures generated through rule base matching are generated using a template-based approach. The specific rectification measures output based on standardized language templates are as follows: The rectification measures adopt the template "In response to the XX problem, the XX department needs to complete the XX rectification within the XX time limit"; Accountability measures use the template "The person responsible for the XX issue will be given XX disciplinary action by the XX department in accordance with regulations"; Preventative measures can follow the template of "To prevent similar problems from happening again, it is necessary to establish the XX mechanism".

7. A system for automatically generating corrective measures based on issues identified during enterprise inspections, characterized in that: The system includes: The problem classification and feature extraction module is used to analyze the lexical, syntactic and semantic structure of inspection problem texts using NLP models, identify problem types, determine the domains involved, assess the severity of problems, and extract keywords. At the same time, it constructs a complete inspection problem labeling system to label problems in multiple dimensions, in the form of: label dimensions including problem type + domain involved + problem severity + keywords. The knowledge graph construction module is used to collect and integrate company rules and regulations as well as historical corporate inspection cases to build a knowledge graph, enabling structured and interconnected storage and efficient retrieval of various types of information; The knowledge graph creation and measure generation module is used to create a rule base based on the knowledge graph and generate candidate rectification measures through "problem tag matching + rule priority sorting + case similarity calculation". The rectification measure generation and optimization module is used to match candidate rectification measures generated through a rule base. It adopts a template-based generation method, outputs standardized rectification measures based on standardized language templates, and optimizes the generated rectification measures by modifying the content, adjusting the priority, adding responsible persons and time limits. At the same time, it combines reinforcement learning algorithms to automatically optimize rule weights based on feedback from historical rectification effects.

8. The automatic generation system for rectification measures based on enterprise inspection problems according to claim 7, characterized in that, Problem types include organizational management issues, operational management issues, and financial management issues; areas include finance, human resources, and general management; severity levels include minor, moderate, and severe; keywords include irregular expense reimbursement, personnel selection and appointment, and lack of oversight. The nodes in the knowledge graph cover key information such as problem type, problem cause, rectification measures, responsible department, and rules and regulations; the relationships between nodes in the knowledge graph include problem → cause, cause → measure, and measure → rule; The rule base includes severity rules, domain rules, and recurrence rules. Among them, severity rules stipulate that if the problem label includes "supply chain disruption risk," then the corrective measures must include "supplier diversification management" and "digital supply chain monitoring." Domain rules stipulate that if the problem involves "financial domain - fund management," then the corrective measures will automatically be associated with "financial audit," "account verification," and "improving the reimbursement system." Recurrence rules stipulate that if similar problems recur in historical cases, then the corrective measures will include "establishing a long-term supervision mechanism" and "interviewing the responsible person." The process of generating candidate rectification measures is as follows: first, the problem is identified and its tags are extracted; then, matching and retrieval are performed in the rule base and knowledge graph; finally, corresponding candidate rectification measures are selected according to rule priority and case similarity. The specific rectification measures based on the standardized language template output specifications are as follows: Rectification measures adopt the template "Regarding the XX problem, the XX department needs to complete the XX rectification within the XX time limit"; accountability measures use the template "The XX department shall give XX punishment to the relevant responsible persons for the XX problem in accordance with discipline and regulations"; and prevention measures adopt the template "In order to avoid similar problems from happening again, it is necessary to establish the XX mechanism".

9. An electronic device, characterized in that, include: Memory and at least one processor; The memory contains computer programs; The at least one processor executes the computer program stored in the memory, causing the at least one processor to perform the automatic generation method for corrective measures based on enterprise inspection issues as described in any one of claims 1 to 6.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that can be executed by a processor to implement the automatic generation method for corrective measures based on enterprise inspection issues as described in any one of claims 1 to 6.

Images