Intelligent data security compliance risk assessment method and system based on structured knowledge base

Abstract

The invention relates to the technical field of data security, and discloses an intelligent data security compliance risk assessment method based on a structured knowledge base, which comprises the following steps of: 1, constructing a dynamic knowledge base, namely performing structured processing on an external strategy file to form assessment key points, executing state locking management on the formed evaluation key points; and step 2, screening and combining evaluation key points in the constructed dynamic knowledge base, and generating a customized inspection scheme containing specific inspection items. According to the technical scheme, a multi-source heterogeneous external strategy file is subjected to structured processing to construct an evaluation key point knowledge base containing a state locking management and dynamic updating mechanism, so that the technical effect of unifying, standardizing and dynamic scattered evaluation basis is achieved; compared with the technical scheme of scattered storage, different structures and manual tracking policy updating in the prior art, the method solves the defects of incomplete assessment reference content coverage and updating lagging.

Description

Technical Field

[0001] This invention relates to the field of data security technology, specifically to an intelligent data security compliance risk assessment method and system based on a structured knowledge base. Background Technology

[0002] With the booming development of the digital economy, data has become a core strategic asset for enterprises. At the same time, global regulations on data security and personal information protection are becoming increasingly stringent, with numerous laws, regulations, and standards being introduced by national and industry authorities. Therefore, conducting regular data security compliance risk assessments has become a necessary step for enterprises to ensure stable business operations and mitigate legal risks.

[0003] Currently, the following technical problems commonly exist in the practice of conducting data security compliance risk assessments:

[0004] Existing technologies face challenges in conducting data security compliance checks due to the dispersed storage and inconsistent structures of assessment criteria such as national regulations and industry standards. There is a lack of effective ways to standardize and integrate multi-source heterogeneous requirements and build a structured knowledge base. At the same time, the lack of a dynamic update mechanism means that the assessment criteria cannot be updated in a timely manner as policies are released, resulting in deficiencies in the coverage and updates of the assessment benchmarks.

[0005] In the process of compliance checks, existing technologies generally adopt manual offline processes. The distribution of assessment plans, the collection and feedback of self-inspection information, and the collaborative verification of multiple parties are all carried out in a non-systematic manner. This model leads to a lack of unified standards in the workflow, resulting in delays and distortions in information transmission, reducing the efficiency of collaborative work, and introducing the risk of operational errors.

[0006] In the analysis of inspection results, existing technologies typically rely on manual data collection and score calculation. Due to the lack of risk quantification scoring rules and calculation models, the generation of assessment results is inefficient, and the scoring conclusions are easily affected by subjective factors. It is not easy to present risk distribution and ranking in an intuitive and visual way, and thus cannot provide effective data support for management decisions.

[0007] In terms of adapting to different inspection needs, existing technologies typically provide tools with built-in fixed assessment processes and inspection items, lacking the ability to flexibly customize and arrange inspection plans as needed. When faced with different business systems, specific industry standards, or special inspection tasks, they cannot quickly generate personalized assessment plans that match them, resulting in insufficient versatility and scenario adaptability of the tools.

[0008] To address these issues, this invention proposes an intelligent data security compliance risk assessment method and system based on a structured knowledge base. Summary of the Invention

[0009] To address the shortcomings of existing technologies, this invention provides an intelligent data security compliance risk assessment method and system based on a structured knowledge base, in order to solve the problems mentioned in the background.

[0010] To achieve the above objectives, the present invention provides the following technical solution: an intelligent data security compliance risk assessment method based on a structured knowledge base, comprising the following steps:

[0011] Step 1: Construct a dynamic knowledge base. This process includes structuring external policy documents to form evaluation points, and performing state locking management on the formed evaluation points.

[0012] Step 2: The evaluation points are selected and combined from the constructed dynamic knowledge base to generate a customized inspection plan containing specific inspection items. The customized inspection plan is then encapsulated, management attributes are added to the plan, and after forming a standardized evaluation task, it is released online.

[0013] Step 3: Receive the self-inspection information submitted for the released standardized assessment task, collect electronic supporting materials, and record operation logs to achieve full-process traceability; conduct online collaborative verification of the collected self-inspection information and electronic supporting materials, and form complete inspection item data with verification conclusions after confirmation by multiple inspectors;

[0014] Step 4: Call up the complete inspection item data and automatically calculate the risk score for each item according to the preset risk calculation model. Then, perform aggregation and ranking processing on all risk scores. Based on the processed risk scores and rankings, automatically generate a structured assessment report with embedded data visualization charts.

[0015] Step 5: Compile the structured evaluation report data generated from each evaluation task; and feed the statistical analysis results back to the dynamic knowledge base to achieve closed-loop optimization of the evaluation system.

[0016] Step five further includes:

[0017] Sub-step The system consists of solidified storage. In the structured evaluation report, each evaluation point in the knowledge base is addressed. Statistics in all Total number of citations in the report Total number of times it was judged as non-compliant Calculate each assessment point according to the following formula. Frequency of non-compliance Frequency of use This forms a statistical data set containing the stated frequencies:

[0018] ,

[0019] ,

[0020] in, For the first in the dynamic knowledge base Key evaluation points, The total number of structured assessment reports. Key points for assessment exist The total number of times it appears in the report, when for hour, for , Key points for assessment exist The total number of times the verification conclusion in the report was non-compliant. Key points for assessment Frequency of non-compliance items Key points for assessment Frequency of use in all evaluation tasks.

[0021] Sub-step The system calls the generated statistical data set, targeting each evaluation point. Based on the frequency of non-compliance items Frequency of use Calculate the optimization adjustment factor If the frequency of use Less than the preset low usage threshold This triggers an archiving suggestion; if the usage frequency is... Not less than the low utilization threshold The optimization adjustment factor The calculation formula is as follows:

[0022] ,

[0023] in, Key points for assessment The calculated weight optimization adjustment values, The preset risk weight adjustment coefficient, Key points for assessment Frequency of non-compliance items The preset average non-compliance rate benchmark value, This is a preset low usage threshold;

[0024] Sub-step The system will calculate the optimization adjustment factor. The archive suggestion, or the one triggered, is fed back to the dynamic knowledge base; if the received factor is the optimization adjustment factor... The system will evaluate the key points. Default risk weight Updated to new risk weights The updated formula is:

[0025] ,

[0026] If the received document is the aforementioned archiving suggestion, the system will automatically send a suggestion to the administrator to review the assessment points. State lock flag Set as The instructions are used to complete the closed-loop optimization of the dynamic knowledge base.

[0027] Preferably, the management of the enabled or disabled status of the evaluation points includes: when an evaluation point is set to the enabled state, locking the data content of the evaluation point so that the data content cannot be modified or deleted;

[0028] The structured processing includes parsing the unstructured text content of external policy documents and storing it in a structured database containing evaluation items, compliance standard details, and typical violation case fields.

[0029] Preferably, the step of publishing the customized inspection plan includes: encapsulating the customized inspection plan into an evaluation task package, wherein the encapsulation includes supplementing the customized inspection plan with inspection cycle and execution requirements information.

[0030] Preferably, receiving the self-inspection information includes: receiving the self-inspection information through an encrypted channel and generating an operation log containing the operator and operation time in real time;

[0031] The inspection item data that forms the verification conclusion includes: allowing multiple inspectors to annotate the self-inspection information online and add verification comments through a collaborative verification interface to form the verification conclusion.

[0032] Preferably, the automatic calculation of risk scores includes: for each inspection item, calculating the individual risk score of the inspection item based on the verification conclusion of the inspection item and the risk weight preset for the inspection item in the knowledge base; and weighting and summing all individual risk scores to obtain the overall risk score.

[0033] The generated assessment report includes: generating a data visualization report containing a risk dashboard and a risk ranking chart based on the calculated risk score.

[0034] Preferably, step one further includes:

[0035] Sub-step It receives unstructured text input from an external policy file and uses a text vectorization algorithm to convert the unstructured text input into an input text vector. And it is consistent with what is already stored in the knowledge base. Text vector of key evaluation points Perform similarity calculations one by one. If all calculated similarity values ​​are less than the preset text similarity threshold... The unstructured text input is determined to be a new evaluation point, and the similarity calculation uses the following cosine similarity formula:

[0036] ,

[0037] in, The feature vector is generated after processing unstructured text input using an algorithm. For the first in the knowledge base The feature vectors corresponding to the text content of the stored evaluation points. This represents the total number of evaluation points already stored in the knowledge base. For the input text vector and the first The dot product of the text vectors has been stored. and For the input text vector and the first The Euclidean norm of the stored text vector. The preset text similarity threshold;

[0038] Sub-step For unstructured text input identified as new assessment points, a structured field extraction operation is performed to generate structured assessment point data containing assessment items, compliance standard details, and typical violation case fields. Furthermore, a default risk weight is calculated based on the risk keywords identified during the structured field extraction operation. The default risk weight It is calculated using the following formula:

[0039] ,

[0040] in, The default risk weight values ​​calculated for the newly added assessment points. The basic risk weights that are applicable to all assessment criteria. To identify the first in the text content The additional weight value corresponding to risk keywords, The total number of identified risk keywords;

[0041] Sub-step In response to an activation command issued by the administrator, the system performs a status activation determination on the generated structured assessment key data. The condition for this status activation determination is that all preset required fields of the structured assessment key data are not empty. When the condition for the status activation determination is met, the system sets a status lock flag for the structured assessment key data. Set as Furthermore, the structured assessment key data is set to an enabled state that is neither modifiable nor deletable, so that it can be called upon in subsequent customized inspection plan generation steps. As a status lock identifier for key data in structured assessment, when for When it is in the enabled state, when for It is currently in an unused state.

[0042] Preferably, step two further includes:

[0043] Sub-step In response to operator input including The system uses the dynamic knowledge base to filter all state lock identifiers in the bit-band weight selection criterion selection instruction. for Structured assessment key data; and for the selected structured assessment key data Calculate key data for structured assessment. The overall matching score with the selection instruction When the comprehensive matching score Greater than the preset matching score threshold At that time, the structured assessment key data will be used. The initial set of inspection items is incorporated into a final customized inspection plan containing specific inspection items, and the comprehensive matching score is used to determine the final inspection plan. The calculation formula is as follows:

[0044] ,

[0045] in, For the first The comprehensive matching score between the structured assessment key data and the selection instructions. The first one selected from the dynamic knowledge base The structured assessment data points are currently in use. The total number of selection criteria included in the selection instruction. For the first Selection criteria, For the first Choose the weighting coefficient corresponding to the standard. To match the score threshold, For the matching determination function, when Field content and selection criteria The function value during matching is When there is a mismatch, it is ;

[0046] Sub-step The system receives the generated customized inspection plan and the management attributes input by the operator, including the inspection cycle and execution requirements; the system encapsulates the customized inspection plan and the management attributes to generate an evaluation task package to be released. Then, a task compliance check is performed, and the condition for passing the task compliance check is the evaluation task package. The start date of the inspection cycle is earlier than the end date, and the number of inspection items included in the customized inspection plan is not less than the preset minimum number. When the compliance verification conditions for the task are met, the system determines that the evaluation task package is valid. For compliant task packages;

[0047] Sub-step For the assessment task package that is determined to be a compliant task package The system assigns globally unique task identifiers. Combined with the globally unique task identifier Solidify this compliance task package to form a standardized assessment task. The system executes the standardized evaluation task. The online publishing operation is determined to be successful through a publishing status confirmation mechanism; if the publishing status confirmation mechanism returns a success flag, the standardized evaluation task is then... The execution status is updated to "Published" so that it can be called in subsequent self-inspection information submission steps.

[0048] Preferably, step three further includes:

[0049] Sub-step The system receives self-inspection information and electronic supporting materials submitted for published standardized assessment tasks, and calculates a data integrity check value for the electronic supporting materials. The system will calculate the data integrity verification value. Original data integrity check value received along with the electronic supporting materials A comparison is performed, and the data integrity verification value is... The original data integrity check value When they are completely equal, the data transmission is deemed complete and error-free. The system then saves the submitted self-inspection information and electronic supporting materials, and generates an operation log containing the operator and operation timestamp.

[0050] Sub-step Regarding the solidified self-inspection information and electronic supporting materials, the system sends... The collaborative verification interface is open to all inspectors; receiving data from each inspector. The score given for the verification comments on this self-inspection information. Combined into the various inspection personnel Preset verification weight coefficient Calculate the comprehensive verification score When the comprehensive verification score is... Reaching the preset verification confidence threshold At that time, the system generates a final verification conclusion, namely the comprehensive verification score. The calculation formula is as follows:

[0051] ,

[0052] in, The overall verification score is obtained by weighting and averaging the opinions of multiple inspectors. The total number of inspectors participating in the collaborative verification. For the first The scores given by the inspectors for their verification opinions. For the first The verification weight coefficient corresponding to each inspector The preset verification confidence threshold;

[0053] Sub-step Based on the calculated comprehensive verification score The system binds the self-inspection information, the electronic supporting materials, and the final verification conclusion to generate complete inspection item data with the verification conclusion. The system provides complete inspection item data with verification conclusions. Mark the status as "Pending Analysis" and include the complete inspection item data with verification conclusions. It is sent to the risk calculation queue for use in subsequent risk score calculation steps.

[0054] Preferably, step four further includes:

[0055] Sub-step The system retrieves complete inspection item data with verification conclusions from the risk calculation queue. The system extracts the final verification conclusion; it then maps this final verification conclusion to a quantified risk coefficient. The system then bases its conclusions on the risk coefficient. Complete inspection item data The corresponding default risk weight in the dynamic knowledge base And the business scenario impact coefficients set for this standardization assessment task. Calculate complete inspection item data Individual risk score The calculation formula is as follows:

[0056] ,

[0057] Here, the risk quantification score is for a single inspection item. This represents the risk coefficient value corresponding to the final verification conclusion. The default risk weight value pre-stored in the knowledge base for the inspection item. This is the risk impact adjustment factor for the inspection item in the specific business scenario of this assessment task;

[0058] Sub-step For all aspects included in a standardized assessment task For each inspection item, the system repeatedly calculates the risk quantification score to obtain a set of individual risk scores. The system performs aggregation operations on the set of individual risk scores according to preset aggregation dimensions to calculate the total aggregated risk score. The system then aggregates the total risk score. The risk level is determined by comparing the risk with preset multi-level risk thresholds. The aggregated risk score The following summation formula is used for the calculation:

[0059] ,

[0060] in, This is the total risk score aggregated according to a specific aggregation dimension. This represents the total number of check items included under the aggregation dimension. For the aggregation dimension, the first Individual risk score for each inspection item;

[0061] The risk level The determination criteria are:

[0062] like > , High risk;

[0063] like ≤ , Medium risk;

[0064] like ≤ , Low risk;

[0065] in, and These are preset thresholds for determining high and low risk.

[0066] Sub-step The system calls the calculated aggregate risk score. With the determined risk level The total aggregation risk score is calculated based on different aggregation dimensions. Sort the data in descending order to generate a risk ranking list;

[0067] The system will compare the risk ranking list with the risk level. The data visualization chart is mapped to a specific style, and a structured evaluation report is automatically generated that embeds the data visualization chart. This structured evaluation report is then stored for subsequent statistical analysis steps.

[0068] An intelligent data security compliance risk assessment system based on a structured knowledge base, the intelligent data security compliance risk assessment system comprising:

[0069] The inspection item knowledge base management module is used to structure external strategy documents to form evaluation points and store them in the knowledge base. The response plan customization instruction selects and combines evaluation points from the knowledge base to generate a customized inspection plan.

[0070] The inspection process management module is used to receive customized inspection plans generated by the inspection item knowledge base management module, process them into evaluation tasks and release them after standardization, and then receive and store self-inspection information and verification conclusions for the evaluation tasks to form complete inspection item data.

[0071] The intelligent result analysis module is used to receive complete inspection item data generated by the inspection process management module, automatically calculate risk scores based on the risk calculation model and generate a visual assessment report, and at the same time, statistically analyze historical assessment data and feed the statistical results back to the inspection item knowledge base management module.

[0072] This invention provides an intelligent data security compliance risk assessment method and system based on a structured knowledge base. It has the following beneficial effects:

[0073] 1. This invention adopts a technical solution that structures multi-source heterogeneous external policy documents into an assessment key knowledge base that includes state locking management and dynamic update mechanism. This achieves the technical effect of unifying, standardizing and dynamizing the dispersed assessment basis. Compared with the existing technical solutions that have dispersed storage of assessment basis, inconsistent structure and reliance on manual tracking of policy updates, this invention solves the shortcomings of incomplete coverage and lagging updates of assessment benchmark content.

[0074] 2. This invention adopts a technical solution that encapsulates the inspection plan into a standardized assessment task and publishes it online. It also supports online submission of self-inspection information and online collaborative verification by multiple parties. This achieves the technical effect of online, automated and standardized assessment of the entire process. Compared with the existing technology that relies on manual offline methods for plan distribution, information collection and collaborative verification, this invention solves the shortcomings of inconsistent workflows, delayed and distorted information transmission and low collaborative efficiency.

[0075] 3. This invention adopts a technical solution that automatically calculates and quantifies risk scores based on a preset risk calculation model and automatically generates a structured assessment report with embedded data visualization charts. This achieves the technical effects of automated inspection result analysis, objective quantification of risk assessment, and intuitive decision support. Compared with existing technologies that rely on manual data collection and score calculation and lack risk quantification rules and visualization methods, this invention solves the shortcomings of low efficiency in generating assessment results, conclusions that are easily influenced by subjectivity, and difficulty in supporting management decisions.

[0076] 4. This invention adopts a technical solution based on a structured knowledge base, which allows users to filter and combine evaluation points as needed to generate customized inspection plans. This achieves the technical effect of flexible configuration and on-demand generation of evaluation plans. Compared with existing technologies that use fixed evaluation processes and inspection items and lack the ability to customize plans, this invention solves the defects of insufficient tool versatility and scenario adaptability when facing diverse inspection needs. Attached Figure Description

[0077] Figure 1 This is a flowchart of the present invention;

[0078] Figure 2 This is a system framework diagram of the present invention. Detailed Implementation

[0079] To enable those skilled in the art to understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort should fall within the scope of protection of the present invention.

[0080] The present invention will now be described in detail with reference to the accompanying drawings:

[0081] Example:

[0082] Please see the appendix Figure 1 and attached Figure 2This invention provides an intelligent data security compliance risk assessment method based on a structured knowledge base, comprising the following steps:

[0083] Step 1: Construct a dynamic knowledge base. This process includes structuring external policy documents to form evaluation points, and performing state locking management on the formed evaluation points.

[0084] Step 2: The evaluation points are selected and combined from the constructed dynamic knowledge base to generate a customized inspection plan containing specific inspection items. The customized inspection plan is then encapsulated, management attributes are added to the plan, and after forming a standardized evaluation task, it is released online.

[0085] Step 3: Receive the self-inspection information submitted for the released standardized assessment task, collect electronic supporting materials, and record operation logs to achieve full-process traceability; conduct online collaborative verification of the collected self-inspection information and electronic supporting materials, and form complete inspection item data with verification conclusions after confirmation by multiple inspectors;

[0086] Step 4: Call up the complete inspection item data and automatically calculate the risk score for each item according to the preset risk calculation model. Then, perform aggregation and ranking processing on all risk scores. Based on the processed risk scores and rankings, automatically generate a structured assessment report with embedded data visualization charts.

[0087] Step 5: Compile the structured evaluation report data generated from each evaluation task; and feed the statistical analysis results back to the dynamic knowledge base to achieve closed-loop optimization of the evaluation system.

[0088] Step one further includes:

[0089] Sub-step It receives unstructured text input from external policy files and uses a text vectorization algorithm to convert the unstructured text input into input text vectors. And it is consistent with what is already stored in the knowledge base. Text vector of key evaluation points Perform similarity calculations one by one. If all calculated similarity values ​​are less than the preset text similarity threshold... Unstructured text input is identified as a new evaluation point, and the similarity is calculated using the following cosine similarity formula:

[0090] ,

[0091] in, The feature vector is generated after processing unstructured text input using an algorithm. For the first in the knowledge base The feature vectors corresponding to the text content of the stored evaluation points. This represents the total number of evaluation points already stored in the knowledge base. For the input text vector and the first The dot product of the text vectors has been stored. and For the input text vector and the first The Euclidean norm of the stored text vector. The preset text similarity threshold;

[0092] Sub-step For unstructured text input identified as new assessment points, a structured field extraction operation is performed to generate structured assessment point data containing assessment items, compliance standard details, and typical violation case fields. Default risk weights are also calculated based on risk keywords identified during the structured field extraction operation. The default risk weight It is calculated using the following formula:

[0093] ,

[0094] in, The default risk weight values ​​calculated for the newly added assessment points. The basic risk weights that are applicable to all assessment criteria. To identify the first in the text content The additional weight value corresponding to risk keywords, The total number of identified risk keywords;

[0095] Sub-step In response to the administrator's activation command, the system performs a status activation determination on the generated structured assessment key data. The condition for this status activation determination is that all the preset required fields of the structured assessment key data are not empty. When the status activation determination condition is met, the system sets a status lock flag for the structured assessment key data. Set as Furthermore, the structured assessment key data is set to an enabled state that is neither modifiable nor deletable, so that it can be called upon in subsequent customized inspection plan generation steps. As a status lock identifier for key data in structured assessment, when for When it is in the enabled state, when for It is currently in an unused state.

[0096] Step two further includes:

[0097] Sub-step In response to operator input including The selection instruction for the bit-band weight selection criterion is used by the system to filter all state lock identifiers from a dynamic knowledge base. for Structured assessment key data; and for the selected structured assessment key data Calculate key data for structured assessment. Overall matching score with selection instructions When the overall matching score Greater than the preset matching score threshold At that time, the structured assessment key data will be used. The initial set of inspection items is incorporated, and a customized inspection plan containing specific inspection items is generated, which is then combined with the overall matching score. The calculation formula is as follows:

[0098] ,

[0099] in, For the first The comprehensive matching score between the structured assessment key data and the selection instructions. The first one selected from the dynamic knowledge base The structured assessment data points are currently in use. The total number of selection criteria included in the selection instruction. For the first Selection criteria, For the first Choose the weighting coefficient corresponding to the standard. To match the score threshold, For the matching determination function, when Field content and selection criteria The function value during matching is When there is a mismatch, it is ;

[0100] Sub-step The system receives the generated customized inspection plan and the management attributes input by the operator, including the inspection cycle and execution requirements. The system then encapsulates the customized inspection plan and management attributes to generate an evaluation task package to be released. Then, a task compliance check is performed. The condition for passing the task compliance check is that the task package is evaluated. The inspection cycle start date is earlier than the end date, and the customized inspection plan includes at least the preset minimum number of inspection items. When the conditions for passing the task compliance verification are met, the system determines that the evaluation task package is valid. For compliant task packages;

[0101] Sub-step Assessment task packages that are determined to be compliant The system assigns globally unique task identifiers. Combined with a globally unique task identifier Solidify this compliance task package to form a standardized assessment task. The system performs standardized evaluation tasks. The online publishing operation is determined to be successful through a publishing status confirmation mechanism; if the publishing status confirmation mechanism returns a success flag, the standardized evaluation task is then completed. The execution status is updated to "Published" so that it can be called in subsequent self-inspection information submission steps.

[0102] Step three further includes:

[0103] Sub-step The system receives self-inspection information and electronic supporting materials submitted for published standardized assessment tasks, and calculates data integrity verification values ​​for the electronic supporting materials. The system will calculate the data integrity verification value. Original data integrity check value received along with electronic supporting materials Perform a comparison; when the data integrity check value... Verification value of the original data integrity When they are completely equal, the data transmission is deemed complete and error-free. The system then saves the submitted self-inspection information and electronic supporting materials, and generates an operation log containing the operator and operation timestamp.

[0104] Sub-step The system provides information to the system based on the solidified self-inspection information and electronic supporting materials. The collaborative verification interface is open to all inspectors; receiving data from each inspector. The score given for the verification comments on this self-inspection information. Combined into the various inspection personnel Preset verification weight coefficient Calculate the comprehensive verification score When the comprehensive verification score Reaching the preset verification confidence threshold At that time, the system generates a final verification conclusion, based on the comprehensive verification score. The calculation formula is as follows:

[0105] ,

[0106] in, The overall verification score is obtained by weighting and averaging the opinions of multiple inspectors. The total number of inspectors participating in the collaborative verification. For the first The scores given by the inspectors for their verification opinions. For the first The verification weight coefficient corresponding to each inspector The preset verification confidence threshold;

[0107] Sub-step Based on the calculated comprehensive verification score The system binds the self-inspection information, electronic supporting materials, and final verification conclusions to generate complete inspection item data with the verification conclusions. The system provides complete inspection item data with verification conclusions. Mark the status as "Pending Analysis" and include the complete inspection item data with verification conclusions. It is sent to the risk calculation queue for use in subsequent risk score calculation steps.

[0108] Step four further includes:

[0109] Sub-step The system retrieves complete inspection item data with verification conclusions from the risk calculation queue. Extract the final verification conclusion; the system maps the final verification conclusion to a quantified conclusion risk coefficient. The system then bases its conclusions on the risk coefficient. Complete inspection item data The corresponding default risk weight in the dynamic knowledge base And the business scenario impact coefficients set for this standardization assessment task. Calculate complete inspection item data Individual risk score The calculation formula is as follows:

[0110] ,

[0111] in, The risk quantification score for a single inspection item, This represents the risk coefficient value corresponding to the final verification conclusion. The default risk weight value pre-stored in the knowledge base for the inspection item. This is the risk impact adjustment factor for the inspection item in the specific business scenario of this assessment task;

[0112] Sub-step For all aspects included in a standardized assessment task For each inspection item, the system repeatedly calculates the risk quantification score to obtain a set of individual risk scores. The system performs aggregation operations on the set of individual risk scores based on preset aggregation dimensions to calculate the total aggregated risk score. The system will then aggregate the total risk score. The risk level is determined by comparing the risk with preset multi-level risk thresholds. Aggregate risk total score The following summation formula is used for the calculation:

[0113] ,

[0114] in, This is the total risk score aggregated according to a specific aggregation dimension. This represents the total number of check items included under the aggregation dimension. For the aggregation dimension, the first Individual risk score for each inspection item;

[0115] Risk level The determination criteria are:

[0116] like > , High risk;

[0117] like ≤ , Medium risk;

[0118] like ≤ , Low risk;

[0119] in, and These are preset thresholds for determining high and low risk.

[0120] Sub-step The system calls to calculate the aggregated risk score. With the determined risk level The total aggregation risk score is calculated based on different aggregation dimensions. Sort the data in descending order to generate a risk ranking list;

[0121] The system will rank the risk list and assign risk levels. By mapping to a specific style of data visualization chart, a structured evaluation report with embedded data visualization chart is automatically generated and stored for subsequent statistical analysis steps.

[0122] Step five further includes:

[0123] Sub-step The system consists of solidified storage. In the structured evaluation report, each evaluation point in the knowledge base is addressed. Statistics in all Total number of citations in the report Total number of times it was judged as non-compliant Calculate each assessment point according to the following formula. Frequency of non-compliance Frequency of use This forms a statistical data set containing frequencies:

[0124] ,

[0125] ,

[0126] in, For the first in the dynamic knowledge base Key evaluation points, The total number of structured assessment reports. Key points for assessment exist The total number of times it appears in the report, when for hour, for , Key points for assessment exist The total number of times the verification conclusion in the report was non-compliant. Key points for assessment Frequency of non-compliance items Key points for assessment Frequency of use in all evaluation tasks.

[0127] Sub-step The system calls the generated statistical data set, targeting each evaluation point. Based on the frequency of non-compliance items Frequency of use Calculate the optimization adjustment factor If frequency is used Less than the preset low usage threshold This triggers an archiving suggestion; if frequency is used... Not less than the low utilization threshold Optimize adjustment factors The calculation formula is as follows:

[0128] ,

[0129] in, Key points for assessment The calculated weight optimization adjustment values, The preset risk weight adjustment coefficient, Key points for assessment Frequency of non-compliance items The preset average non-compliance rate benchmark value, This is a preset low usage threshold;

[0130] Sub-step The system will calculate the optimization adjustment factor. Or, if an archiving suggestion is triggered, it will be fed back to the dynamic knowledge base; if the received factor is an optimization adjustment factor... The system will evaluate the key points. Default risk weight Updated to new risk weights The updated formula is:

[0131] ,

[0132] If the received suggestion is for archiving, the system will automatically send a suggestion to the administrator to include the key evaluation points. State lock flag Set as The instructions are used to complete the closed-loop optimization of the dynamic knowledge base.

[0133] Step one fundamentally solves the problems of incomplete coverage and delayed updates of assessment criteria by transforming the scattered and heterogeneous strategy documents into structured, machine-readable assessment points with dynamic update capabilities.

[0134] Step two enables the flexible and efficient transformation of assessment solutions from knowledge bases to executable tasks, allowing for the rapid customization and issuance of standardized assessment tasks based on specific business scenarios. This addresses the shortcomings of traditional inspection solutions, which are rigid and unable to adapt to diverse assessment needs, and significantly improves the targeting and efficiency of assessment task initiation.

[0135] Step 3 fully digitizes and standardizes the collection and verification process of self-inspection information, ensures data reliability through data verification and log traceability, and enhances the objectivity of verification conclusions through a multi-party collaborative verification mechanism, thereby solving the problems of distorted information transmission, low collaborative efficiency, and difficulty in tracing the process in offline processes.

[0136] Step four involves introducing a risk calculation model to automate and quantify the assessment results, delegating the tedious analysis work to the system, ensuring the objectivity and accuracy of the scoring, and presenting the risk situation intuitively through a visual report, thus addressing the pain points of low efficiency, subjective conclusions, and difficulty in supporting decision-making in manual analysis.

[0137] Step 5 involves constructing a data feedback and self-optimization closed loop from assessment results to a knowledge base. This enables the assessment system to learn and evolve continuously based on historical data, solving the problems of rigidity and inability to dynamically adapt to changes in risk in traditional assessment systems, and ensuring the long-term effectiveness and accuracy of the assessment model.

[0138] An intelligent data security compliance risk assessment system based on a structured knowledge base, comprising:

[0139] The inspection item knowledge base management module is used to structure external policy documents to form evaluation points and store them in the knowledge base. The response plan customization instruction selects and combines evaluation points from the knowledge base to generate a customized inspection plan.

[0140] The inspection process management module is used to receive customized inspection plans generated by the inspection item knowledge base management module, process them into evaluation tasks and release them after standardization, and then receive and store self-inspection information and verification conclusions for the evaluation tasks to form complete inspection item data.

[0141] The intelligent results analysis module receives complete inspection item data generated by the inspection process management module, automatically calculates risk scores based on the risk calculation model, generates a visual assessment report, and simultaneously compiles historical assessment data and feeds the statistical results back to the inspection item knowledge base management module.

[0142] The inspection item knowledge base management module ensures that the source data for all assessment activities is authoritative and standardized through structured management and dynamic maintenance of compliance evidence. This provides a solid foundation for generating high-quality, highly relevant, and customized inspection solutions, thereby enhancing the professionalism of the assessment system.

[0143] The inspection process management module enables online and process-oriented control of the entire lifecycle of assessment tasks, greatly improving the efficiency, collaboration, and standardization of process data in assessment work.

[0144] The intelligent results analysis module transforms complex inspection data into intuitive risk insights through automated risk calculation and visualized results presentation. It also drives the self-optimization of the knowledge base, achieving an intelligent leap from data to decision-making and ensuring the continuous advancement of the assessment system.

[0145] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for intelligent data security compliance risk assessment based on structured knowledge base, characterized in that, Comprising the following steps: Step one, constructing a dynamic knowledge base, which includes structuring external policy files to form evaluation points, and performing state locking management on the formed evaluation points; Step two, screening and combining evaluation points from the constructed dynamic knowledge base to generate a customized inspection scheme containing specific inspection items, packaging the customized inspection scheme, supplementing the scheme with management attributes, and forming a standardized evaluation task before performing online release; Step three, receiving self-check information submitted by the released standardized evaluation task, collecting electronic evidence materials, and recording operation logs to realize full-process tracing; open online collaborative verification of the collected self-check information and electronic evidence materials, and form complete inspection item data with verification conclusions through multi-party inspection personnel confirmation; Step four, calling the formed complete inspection item data, automatically calculating the risk score of each item according to the preset risk calculation model, and performing aggregation and ranking processing on all risk scores; Based on the processed risk scores and rankings, a structured evaluation report is automatically generated, which embeds data visualization charts; Step five, statistics of structured evaluation report data generated by previous evaluation tasks; And the statistical analysis results are fed back to the dynamic knowledge base to realize closed-loop optimization of the evaluation system; The step five further comprises: sub-step , the system stores the structured assessment reports in a knowledge base , for each assessment point in the knowledge base , the total number of times cited in all reports , the total number of times judged as not meeting the requirements , the frequency of non-compliance items and the frequency of use of each assessment point are calculated according to the following formula , and a statistical data set containing the frequencies is formed ， ， wherein, is the number of the first evaluation point, is the total number of parts of the structured evaluation report, is the evaluation point occurs in the total number of reports, when is , is , is the evaluation point occurs in the total number of reports with a conclusion of non-conformity, is the evaluation point of the frequency of non-conformities, is the evaluation point of the frequency of use in all evaluation tasks. sub-step , the statistical data set generated by system call, for each evaluation point , according to the frequency of non-compliance and the frequency of use Calculate the optimization adjustment factor ; if the frequency of use is less than the preset low use threshold , trigger the archiving suggestion; if the frequency of use is not less than the low use threshold , the calculation formula of the optimization adjustment factor is as follows: ， wherein, evaluation points calculated weight optimization adjustment value, preset risk weight adjustment coefficient, evaluation points nonconformance occurrence frequency, preset average nonconformance rate baseline value, preset low usage rate threshold value; Sub-step The system will feed back the calculated optimization adjustment factor or the triggered archiving suggestion to the dynamic knowledge base; if the received is the optimization adjustment factor , the system will update the default risk weight of the evaluation key point to the new risk weight , and the update formula is: ， If the received is the archiving suggestion, the system automatically sends an instruction to the administrator to lock the status of the assessment point to the state of the lock to complete the closed-loop optimization of the dynamic knowledge base.

2. The method of claim 1, wherein the method is based on a structured knowledge base of intelligent data security compliance risk assessment. The state management of the evaluation points includes: when the evaluation points are set to an enabled state, the data content of the evaluation points is locked, and the data content cannot be modified or deleted; The structured processing includes: parsing and storing the unstructured text content of the external policy file into a structured database containing evaluation items, compliance standard details, and typical violation case fields.

3. The method of claim 1, wherein the method further comprises: The release of the customized inspection scheme includes: packaging the customized inspection scheme into an evaluation task package, which includes supplementing the customized inspection scheme with inspection period and execution requirement information.

4. The method of claim 1, wherein the method further comprises: The reception of self-check information includes: receiving the self-check information through an encrypted channel, and generating operation logs containing operation personnel and operation time in real time; The formation of inspection item data with verification conclusions includes: allowing multiple inspectors to add verification opinions to the self-check information through a collaborative verification interface to form the verification conclusions.

5. The method of claim 1, wherein the method further comprises: The automatic calculation of risk scores includes: for each inspection item, based on the verification conclusion of the inspection item and the risk weight preset for the inspection item in the knowledge base, the single risk score of the inspection item is calculated; all single risk scores are weighted and summarized to obtain the total risk score; The generation of the evaluation report includes: based on the calculated risk scores, a data visualization report containing a risk dashboard and a risk ranking chart is generated.

6. The method of claim 1, wherein the method further comprises: In the step one, further comprising: Sub-step It receives unstructured text input from an external policy file and uses a text vectorization algorithm to convert the unstructured text input into an input text vector. And it is consistent with what is already stored in the knowledge base. Text vector of key evaluation points Perform similarity calculations one by one. If all calculated similarity values ​​are less than the preset text similarity threshold... The unstructured text input is determined to be a new evaluation point, and the similarity calculation uses the following cosine similarity formula: ， in, The feature vector is generated after processing unstructured text input using an algorithm. For the first in the knowledge base The feature vectors corresponding to the text content of the stored evaluation points. This represents the total number of evaluation points already stored in the knowledge base. For the input text vector and the first The dot product of the text vectors has been stored. and For the input text vector and the first The Euclidean norm of the stored text vector. The preset text similarity threshold; sub-step , for the unstructured text input determined as a new evaluation point, performing a structured field extraction operation to generate structured evaluation point data containing the evaluation item, compliance standard details, and typical violation case fields, and calculating a default risk weight according to the risk keywords identified in the structured field extraction operation , the default risk weight is calculated by the following formula: ， wherein, a default risk weight value calculated for the newly added evaluation point, a base risk weight value possessed by all evaluation points, an additional weight value corresponding to the i-th risk keyword identified in the text content, a total number of the identified risk keywords, a total number of the identified risk keywords. Sub-step , in response to the enabling instruction issued by the administrator, performing a state enabling determination on the generated structured evaluation point data, a condition of the state enabling determination being that all preset mandatory fields of the structured evaluation point data are non-empty, when the condition of the state enabling determination is met, the system locks the state identifier of the structured evaluation point data is set to , and the structured evaluation point data is set to an enabling state of being non-modifiable and non-deletable for subsequent customized inspection scheme generation step calling, wherein, is the state lock identifier of the structured evaluation point data, when is , the enabling state, when is , the deprecated state.

7. The method of claim 1, wherein the method further comprises: In the step two, further comprising: Sub-step In response to operator input including The system uses the dynamic knowledge base to filter all state lock identifiers in the bit-band weight selection criterion selection instruction. for Structured assessment key data; and for the selected structured assessment key data Calculate key data for structured assessment. The overall matching score with the selection instruction When the comprehensive matching score Greater than the preset matching score threshold At that time, the structured assessment key data will be used. The initial set of inspection items is incorporated into a final customized inspection plan containing specific inspection items, and the comprehensive matching score is used to determine the final inspection plan. The calculation formula is as follows: ， wherein, is the number of the first structured assessment point data, is the comprehensive matching score of the structured assessment point data and the selection instruction, is the structured assessment point data in the enabled state selected from the dynamic knowledge base, is the total number of selection criteria contained in the selection instruction, is the first selection criterion, is the first weight coefficient corresponding to the selection criterion, is the matching score threshold value, is the matching decision function, when the field content of the first selection criterion matches the selection criterion , the function value is , and when it does not match, the function value is . Sub-step , receiving the generated customized inspection scheme and receiving the management attributes including the inspection period and execution requirements input by the operator; the system encapsulates the customized inspection scheme and the management attributes to generate an evaluation task package to be released ; and performing a task compliance check after the execution of the task, the passing condition of the task compliance check being that the starting date of the inspection period in the evaluation task package is earlier than the deadline, and the number of inspection items included in the customized inspection scheme is not less than the preset minimum number lower limit ; when the passing condition of the task compliance check is satisfied, the system determines that the evaluation task package is a compliant task package; sub-step , for the evaluation task package determined as a compliance task package , the system assigns a globally unique task identifier , in combination with the globally unique task identifier , the compliance task package is solidified to form a standardized evaluation task ; the system executes the standardized evaluation task ; the system performs an online publishing operation, determines whether the online publishing operation is successful through a publishing state confirmation mechanism; if the publishing state confirmation mechanism returns a success identifier, the execution state of the standardized evaluation task is updated to "published" for subsequent self-check information submission steps.

8. The method of claim 1, wherein the method further comprises: In the step three, further comprising: Sub-step The system receives the self-check information and electronic evidence materials submitted for the published standardized evaluation task, and calculates a data integrity check value for the electronic evidence materials ; The system compares the calculated data integrity check value with the original data integrity check value received together with the electronic evidence materials, and when the data integrity check value is completely equal to the original data integrity check value , it is determined that the data transmission is complete and correct, the system solidifies the submitted self-check information and electronic evidence materials, and generates an operation log containing the operator and operation time stamp; Sub-step Regarding the solidified self-inspection information and electronic supporting materials, the system sends... The collaborative verification interface is open to all inspectors; receiving data from each inspector. The score given for the verification comments on this self-inspection information. Combined into the various inspection personnel Preset verification weight coefficient Calculate the comprehensive verification score When the comprehensive verification score is... Reaching the preset verification confidence threshold At that time, the system generates a final verification conclusion, namely the comprehensive verification score. The calculation formula is as follows: ， in, The overall verification score is obtained by weighting and averaging the opinions of multiple inspectors. The total number of inspectors participating in the collaborative verification. For the first The scores given by the inspectors for their verification opinions. For the first The verification weight coefficient corresponding to each inspector The preset verification confidence threshold; Sub-step , based on the calculated comprehensive review score With the final review conclusion, the system data binds the self-check information, the electronic supporting materials and the final review conclusion to generate complete inspection item data with review conclusion ; The system marks the complete inspection item data with review conclusion The state of "to be analyzed", and transfer the complete inspection item data with review conclusion To the risk calculation queue for subsequent risk score calculation steps.

9. The method of claim 1, wherein the method further comprises: In the step four, further comprising: Sub-step , the system calls the complete check item data with the verification conclusion from the risk calculation queue , extracts the final verification conclusion; the system maps the final verification conclusion into a quantitative conclusion risk coefficient ; then the system calculates the single risk score of the complete check item data according to the conclusion risk coefficient , the corresponding default risk weight of the complete check item data , and the business scene influence coefficient set for this standardized evaluation task , the calculation formula is as follows:​​​ ， wherein, is the risk quantification score of a single inspection item, is the risk coefficient value corresponding to the final verification conclusion, is the pre-stored default risk weight value of the inspection item in the knowledge base, is the risk impact adjustment coefficient of the inspection item in the specific business scenario of the current evaluation task; Sub-step , for all items contained in a standardization evaluation task , the system repeatedly performs the calculation of the risk quantification score to obtain a set of single-item risk scores ; the system performs an aggregation operation on the set of single-item risk scores according to a preset aggregation dimension to calculate an aggregated risk total score ; then the system compares the aggregated risk total score with a preset multi-level risk threshold to determine the risk level , the calculation of the aggregated risk total score uses the following summation formula: ， wherein, is the total risk score after aggregation by a certain aggregation dimension, is the total number of inspection items contained under the aggregation dimension, is the single risk score of the inspection item under the aggregation dimension; the risk level the determination condition is that If > , High risk; If ≤ , is medium risk; If ≤ , is low risk; wherein, with are preset high and low risk decision thresholds; Sub-step , the system calls the calculated aggregate risk total score With the determined risk level , the aggregate risk total score of different aggregate dimensions In descending order, generate a risk ranking list; The system will compare the risk ranking list with the risk level. The data visualization chart is mapped to a specific style, and a structured evaluation report is automatically generated that embeds the data visualization chart. This structured evaluation report is then stored for subsequent statistical analysis steps.

10. An intelligent data security compliance risk assessment system based on structured knowledge base, according to the intelligent data security compliance risk assessment method based on structured knowledge base in any one of claims 1 to 9, characterized in that, The intelligent data security compliance risk assessment system comprises: The check item knowledge base management module is configured to form evaluation points by structurally processing external policy files, store the evaluation points in a knowledge base, and generate a customized check scheme by screening and combining the evaluation points from the knowledge base in response to a scheme customization instruction; The check process management module is configured to receive the customized check scheme generated by the check item knowledge base management module, form evaluation tasks by standardizing the customized check scheme, and publish the evaluation tasks, and then receive and store self-check information and verification conclusions for the evaluation tasks to form complete check item data; The intelligent result analysis module is configured to receive the complete check item data formed by the check process management module, automatically calculate a risk score based on a risk calculation model, and generate a visual evaluation report, and simultaneously statistically process historical evaluation data and feed back statistical results to the check item knowledge base management module.

Images