A data processing method for compliance assessment of data element processing operations

Abstract

The present invention discloses a data processing method for compliance evaluation of data element processing operations, which relates to the technical fields of data processing and compliance management. It includes collecting data elements and regulatory rules and performing preprocessing. Based on the preprocessed data elements and regulatory rules, a knowledge graph associated with data elements and regulatory rules is constructed. The data element processing operations are monitored in real time, operation meta-information is extracted, rule matching is performed in combination with the knowledge graph, the rule matching results, the knowledge graph, and the operation meta-information are fused to obtain data processing features. A compliance evaluation model is constructed using machine learning algorithms. The data processing features are input into the compliance evaluation model to analyze potential risks, and early warnings and restrictions are imposed on high-risk operations. The compliance evaluation model is optimized according to user feedback, constituting a dynamic compliance evaluation method, which is of great significance to the fields of data processing and compliance management.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing and compliance management, and in particular to a data processing method for compliance evaluation of data element processing operations. Background Art

[0002] With the rapid development of information technology and the wide popularization of big data applications, the processing and circulation of data elements have become an important part of modern social and enterprise operations. Data processing operations often involve the use and sharing of sensitive data such as personal privacy and business secrets. Regulatory agencies in various countries have successively introduced a series of laws and regulations related to data protection. The implementation of these laws and regulations provides clear guidelines for the legitimate use of data elements, but also poses increasingly high compliance requirements for data processing activities. In the prior art, the compliance of data processing operations is usually evaluated through manual review, rule matching tools, and simple automation tools. With the increase in data scale and complexity, traditional methods gradually expose problems such as low efficiency, insufficient accuracy, and difficulty in dynamically adapting to regulatory changes, and cannot meet the complex needs of modern enterprises in data compliance management.

[0003] There are mainly many deficiencies in the prior art in the field of data compliance evaluation. Most existing compliance evaluation models lack a dynamic feedback mechanism and are difficult to optimize the compliance evaluation model according to user feedback and changes in the operating environment, resulting in a gradual decline in the prediction ability of the compliance evaluation model. The real-time requirement of data processing operations is increasing day by day, while the prior art often performs poorly in real-time monitoring and dynamic risk assessment, and cannot efficiently capture operation meta-information and conduct rapid evaluation. The impact of data processing operation factors on compliance evaluation has not been fully reflected in existing compliance evaluation models, resulting in evaluation results that are difficult to comprehensively reflect the actual risks of data processing operations. Against this background, how to effectively combine regulatory knowledge graphs, real-time data monitoring, and dynamic risk assessment in the prior art has become a hot issue of common concern in the academic and industrial fields. Summary of the Invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides a data processing method for compliance evaluation of data element processing operations to solve the problem that the real-time requirement of data processing operations is increasing day by day, while the prior art often performs poorly in real-time monitoring and dynamic risk assessment, and cannot efficiently capture operation meta-information and conduct rapid evaluation.

[0006] To solve the above technical problems, the present invention provides the following technical solutions:

[0007] In a first aspect, the present invention provides a data processing method for compliance evaluation of data element processing operations, which includes,

[0008] Collect data elements and regulatory rules, and perform preprocessing;

[0009] Based on the preprocessed data elements and regulatory rules, construct a knowledge graph that associates data elements with regulatory rules;

[0010] Monitor the data element processing operations in real time, extract operation meta-information, and perform rule matching in combination with the knowledge graph;

[0011] Fuse the rule matching results, the knowledge graph, and the operation meta-information to obtain data processing features;

[0012] Use machine learning algorithms to construct a compliance evaluation model, input the data processing features into the compliance evaluation model, analyze potential risks, and give early warnings and restrictions on high-risk operations;

[0013] Collect user feedback and optimize the compliance evaluation model according to the user feedback; The specific steps for constructing the knowledge graph that associates data elements with regulatory rules are as follows:

[0014] Calculate the data elements and the specific conditions in the regulatory rules for semantic similarity , and the expression is:

[0015] ;

[0016] Calculate the association degree between data elements and regulatory rules, and the expression is:

[0017] ;

[0018] Among them, represents the data element, represents the regulatory rule, represents the association degree, represents the cosine similarity, represents the regulatory rule in the specific conditions, represents the data element and the specific conditions in the regulatory rule for semantic similarity, represents the weight of the regulatory rule , represents the complexity of the data element , represents the complexity of the regulatory rule .

[0019] As a preferred solution of the data processing method for compliance evaluation of data element processing operations according to the present invention, wherein: collecting data elements and regulatory rules and performing preprocessing, the specific steps are as follows:

[0020] The data elements refer to user data, device data, and transaction data;

[0021] Performing preprocessing on the data elements means data cleaning, data standardization, and data annotation;

[0022] Performing preprocessing on the regulatory rules means using natural language processing NLP for text parsing and extraction to obtain specific condition entries, structured storage, and rule conflict detection.

[0023] As a preferred solution of the data processing method for compliance evaluation of data element processing operations according to the present invention, wherein: performing real-time monitoring on the data element processing operations, extracting operation meta-information, and performing rule matching in combination with a knowledge graph, the specific steps are as follows:

[0024] Listening to the event stream of data processing operations through the event-driven monitoring tool Kafka to capture operation meta-information , wherein, represents the data elements involved in the operation, represents the operation frequency;

[0025] Performing rule matching in combination with a knowledge graph, the expression is:

[0026] ;

[0027] wherein, represents the matching degree, represents the data elements involved in the operation and the regulatory rules in the specific conditions of the semantic similarity, represents the natural number base, represents the adjustment coefficient, represents the operation frequency attenuation factor.

[0028] As a preferred solution of the data processing method for compliance evaluation of data element processing operations according to the present invention, wherein: fusing the rule matching result, the knowledge graph, and the operation meta-information to obtain data processing features, the expression is:

[0029] ;

[0030] wherein, represents the data processing features, represents the complexity of the data elements involved in the operation, The adjustment parameter representing the matching degree, The adjustment parameter representing the operation frequency, Represents the natural logarithm function.

[0031] As a preferred solution of the data processing method for compliance evaluation of the data element processing operation described in the present invention, wherein: the steps of constructing the compliance evaluation model using the machine learning algorithm are as follows:

[0032] Select the deep neural network DNN algorithm to construct the compliance evaluation model;

[0033] Set the architecture of the compliance evaluation model according to the deep neural network DNN;

[0034] The architecture of the compliance evaluation model refers to the input layer, hidden layer, and output layer;

[0035] Collect a large number of historical compliance evaluation data sets;

[0036] Divide the historical compliance evaluation data set into a training set, a test set, and a validation set according to a ratio, and input them into the compliance evaluation model to train, test, and validate the compliance evaluation model respectively.

[0037] As a preferred solution of the data processing method for compliance evaluation of the data element processing operation described in the present invention, wherein: the steps of inputting the data processing features into the compliance evaluation model, analyzing potential risks, and warning and restricting high-risk operations are as follows:

[0038] Input the data processing features obtained by fusing the rule matching result, the knowledge graph, and the operation meta-information into the compliance evaluation model, and calculate the potential risk value. The expression is:

[0039] ;

[0040] Wherein, Represents the potential risk value, Represents The activation function, Represents the weight matrix, Represents the transpose operation, Represents the bias term;

[0041] Set the threshold , when It means that the current operation is a low-risk operation and no processing is performed;

[0042] When It means that the current operation is a high-risk operation, prevent the current operation from proceeding, and send a warning to the user.

[0043] As a preferred solution of the data processing method for compliance evaluation of the data element processing operation described in the present invention, wherein: the steps of collecting user feedback and optimizing the compliance evaluation model according to the user feedback are as follows:

[0044] Based on the user feedback, calculate the data processing feature offset , and the expression is:

[0045] ;

[0046] Wherein, represents the number of samples of user feedback, represents the index of the sample of user feedback, represents the feedback value marked by the user of the th sample of user feedback, represents the potential risk value of the th sample of user feedback, represents the data processing feature of the th sample of user feedback;

[0047] Add a correction term of the feedback feature offset to the input layer of the compliance evaluation model, and redefine the output formula of the risk value. The expression is:

[0048] ;

[0049] represents the adjusted risk value, represents the feature offset 's adjustment factor.

[0050] In a second aspect, the present invention provides a computer device, including a memory and a processor. The memory stores a computer program, wherein: when the computer program is executed by the processor, it implements any step of the data processing method for compliance evaluation of the data element processing operation described in the first aspect of the present invention.

[0051] In a third aspect, the present invention provides a computer-readable storage medium, on which a computer program is stored, wherein: when the computer program is executed by the processor, it implements any step of the data processing method for compliance evaluation of the data element processing operation described in the first aspect of the present invention.

[0052] The beneficial effects of the present invention are as follows: Data query and preprocessing solve the problems of inconsistent data and regulatory rule sources and low quality, providing high-quality input data for the entire process. Knowledge graph construction enhances the accuracy and intelligence of rule matching through quantitative correlation and semantic matching. Real-time monitoring and rule matching combine real-time monitoring and dynamic frequency attenuation, improving the accuracy and timeliness of evaluation. Data processing feature extraction generates data features that comprehensively reflect risks through multi-dimensional fusion, providing reliable and efficient input for the compliance evaluation model. Through the key steps of data query preprocessing, knowledge graph construction, real-time monitoring and rule matching, feature extraction, and user feedback optimization, the present invention effectively solves the problems of inaccurate rule matching, lack of dynamic adaptability, and insufficient real-time performance in the prior art, constituting a dynamic, accurate, and efficient compliance evaluation method, which is of great significance to the fields of data processing and compliance management. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] To more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0054] Figure 1 It is a flowchart of the data processing method for compliance evaluation of data element processing operations in Embodiment 1.

[0055] Figure 2 It is a schematic diagram for calculating the potential risk value in Embodiment 1. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0056] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following will provide a detailed description of the specific embodiments of the present invention with reference to the drawings in the specification.

[0057] In the following description, many specific details are set forth to fully understand the present invention. However, the present invention can also be implemented in other ways different from those described herein. Those skilled in the art can make similar extensions without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0058] Secondly, the so-called "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that can be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment that is mutually exclusive with other embodiments.

[0059] Embodiment 1, refer to Figure 1 andFigure 2 , which is the first embodiment of the present invention. This embodiment provides a data processing method for compliance evaluation of data element processing operations, including the following steps:

[0060] S1. Collect data elements and regulatory rules, and perform preprocessing;

[0061] The data query tool refers to the SQL query tool and the API interface;

[0062] The database refers to the enterprise internal database and the regulatory database;

[0063] The data elements refer to user data, device data, and transaction data;

[0064] The regulatory rules refer to the texts of laws and regulations related to data elements;

[0065] Performing preprocessing on data elements means data cleaning, data standardization, and data annotation;

[0066] Use the pandas tool to complete missing value processing, outlier detection, duplicate value removal, and format unification;

[0067] Use the normalization, standardization, and encoding tools provided by sklearn.preprocessing to complete the standardization of numerical and categorical data;

[0068] Use Label Studio to implement data annotation;

[0069] Performing preprocessing on regulatory rules means using natural language processing (NLP) for text parsing and extraction to obtain specific conditions, structured storage, and rule conflict detection;

[0070] Implement structured storage through the relational database MySQL;

[0071] Use the logic engine Prolog for rule conflict detection;

[0072] The data query tool collects data elements and regulatory rules from the enterprise internal database and the regulatory database, combined with data cleaning, standardization, and annotation, significantly improving data quality and solving the problems of inconsistent data sources and non-standard formats in the prior art;

[0073] Collect data elements and regulatory rules from the enterprise internal database and the regulatory database through the SQL query tool and the API interface, and perform preprocessing on the data, including data cleaning, standardization, and annotation; at the same time, perform text parsing and extraction, structured storage, and rule conflict detection on the regulatory rules;

[0074] This step resolves the inconsistency issues in the format and storage form of data elements from different sources, such as user data, device data, transaction data, and regulatory rules (legal texts), through a unified query tool and interface. The preprocessing process further cleans and standardizes the data and regulatory rules to ensure data consistency and structured storage, thereby providing high-quality input data for subsequent knowledge graph construction. In addition, rule conflict detection helps to identify and resolve contradictions between different regulatory rules, avoiding incorrect matches caused by inconsistent rules;

[0075] The targeted preprocessing of text parsing and extraction, structured storage, and rule conflict detection ensures that regulatory rules can be efficiently stored and invoked in a structured form, making the subsequent knowledge graph construction more accurate;

[0076] Through this step, the quality and standardization of data and regulatory rules can be significantly improved, providing a solid foundation for the subsequent association construction of the knowledge graph. Compared with the inefficient mode of relying on manual data collection and processing in the existing technology, this step greatly improves the processing efficiency and accuracy, ensuring the reliability of the compliance assessment process;

[0077] This step provides high-quality basic data for knowledge graph construction and rule matching through standardized data and regulatory inputs, thereby improving the accuracy and reliability of the entire compliance assessment process.

[0078] S2. Based on the preprocessed data elements and regulatory rules, construct a knowledge graph that associates data elements with regulatory rules;

[0079] Use the cosine similarity function to calculate the data elements and the specific conditions in the regulatory rules for semantic similarity , and the expression is:

[0080] ;

[0081] Calculate the association degree between data elements and regulatory rules, and the expression is:

[0082] ;

[0083] Among them, represents the data element, represents the regulatory rule, represents the association degree, represents the cosine similarity, represents the regulatory rule and the specific conditions in it, represents the data element and the regulatory rule and the specific conditions in it Semantic similarity indicating the weight of regulatory rules indicating the complexity of data elements indicating the complexity of regulatory rules ;

[0084] Construct a knowledge graph based on pre - processed data elements and regulatory rules, and use the correlation degree to solve the problem in traditional technologies that the matching of data and regulatory rules only depends on keywords and ignores semantic relevance;

[0085] Through the semantic similarity between data elements and the specific conditions in regulatory rules introduced, combined with the regulatory weight , a strategy of preferentially matching important regulations is realized, enhancing the accuracy of rule matching;

[0086] The knowledge graph expresses the complex relationship between data elements and regulatory rules in a graph structure, facilitating the rapid invocation of associated information in real - time monitoring for efficient rule matching;

[0087] Through the construction of the knowledge graph, the complex relationship between data elements and regulatory rules can be expressed in a graph structure. This structured expression method greatly enhances the flexibility and accuracy of subsequent rule matching. The semantic similarity calculation in the formula solves the problem that traditional keyword matching cannot capture deep semantic relationships, and through the introduction of regulatory weight and normalization processing of complexity, the priority of important rules and the fairness of matching are further optimized;

[0088] This step can not only accurately reflect the semantic association between data elements and regulatory rules, but also improve the efficiency of rule matching through the structured storage of the knowledge graph. Compared with the simple string matching method in the existing technology, this step makes rule matching more intelligent and semantic, and provides a basis for rapid invocation in real - time monitoring.

[0089] S3. Conduct real - time monitoring on data element processing operations, extract operation meta - information, and perform rule matching in combination with the knowledge graph;

[0090] Use the event - driven monitoring tool Kafka to listen to the event stream of data processing operations and capture operation meta - information , where represents the data elements involved in the operation represents the operation frequency;

[0091] ​​Perform rule matching in combination with a knowledge graph, and the expression is:

[0092] ;

[0093] Among them, represents the matching degree, represents the data elements involved in the operation and the specific conditions in the regulatory rules semantic similarity, represents the natural logarithm base, represents the adjustment coefficient, represents the operation frequency attenuation factor;

[0094] Use the Kafka event-driven tool to monitor data processing operations in real time and capture operation meta-information , solving the problem of lack of real-time performance in the prior art;

[0095] The rule matching formula dynamically combines the operation frequency and semantic similarity, avoiding misjudgment of high-frequency and low-risk operations and improving the robustness of the matching;

[0096] The introduction of the real-time monitoring tool enables the evaluation of data processing operations to capture the meta-information of data elements and operation frequency in real time, solving the problem of insufficient real-time performance requirements in the prior art. By combining the knowledge graph with the frequency attenuation factor in the formula , dynamically balances the impact of high-frequency operations on rule matching, effectively avoiding the situation where high-frequency and low-risk operations are misjudged as high-risk;

[0097] Through the combination of real-time monitoring and dynamic rule matching, it can quickly capture the characteristics of data operations and accurately evaluate their compliance. This method not only improves the real-time performance, but also improves the robustness of rule matching through the design of the frequency attenuation factor, avoiding misjudgment caused by high-frequency operations in the prior art;

[0098] The combination of real-time monitoring and rule matching not only supports dynamic risk assessment, but also provides key input data for subsequent feature extraction.

[0099] S4. Integrate the rule matching result, knowledge graph and operation meta-information to obtain data processing features;

[0100] The data processing feature expression is:

[0101] ;

[0102] Among them, represents the data processing feature, represents the complexity of the data elements involved in the operation, represents the adjustment parameter of the matching degree, An adjustment parameter representing the operating frequency represents the natural logarithm function;

[0103] The data processing feature formula combines multi-dimensional factors such as matching degree, correlation degree, semantic similarity, and frequency attenuation, solving the limitations of single features in the prior art that cannot comprehensively reflect risks;

[0104] This step generates a feature value with high expression ability by integrating multi-dimensional features of data processing operations , providing reliable support for the input of the subsequent compliance evaluation model. In the formula, the natural logarithm function highlights the influence of the matching degree on the feature, while the frequency attenuation factor and the complexity normalization term effectively balance the contributions of different features to the final feature value;

[0105] The introduction of the natural logarithm function and the frequency attenuation factor can more precisely balance the non-linear relationship between multi-factors, ensuring the academic and practical nature of feature extraction;

[0106] As the core feature for the input of the compliance evaluation model, the accurate extraction of the data processing feature directly determines the accuracy of risk assessment;

[0107] By fusing the rule matching result and the operation meta-information, this step generates data features that can comprehensively reflect the risks of data processing operations. This feature extraction method is more comprehensive than the prior art, solving the problem that single features cannot accurately describe operation risks, and providing guarantee for the efficient learning of the subsequent compliance evaluation model.

[0108] S5. Use machine learning algorithms to construct a compliance evaluation model, input the data processing features into the compliance evaluation model, analyze potential risks, and give early warnings and restrictions on high-risk operations;

[0109] Select the deep neural network DNN algorithm to construct the compliance evaluation model;

[0110] Set the architecture of the compliance evaluation model according to the deep neural network DNN;

[0111] The architecture of the compliance evaluation model refers to the input layer, hidden layer, and output layer;

[0112] Collect a large number of historical compliance evaluation data sets;

[0113] Divide the historical compliance evaluation data sets into training sets, test sets, and validation sets according to a certain proportion, and input them into the compliance evaluation model to train, test, and validate the compliance evaluation model respectively;

[0114] The compliance assessment model is trained using the binary cross - entropy loss function, and the expression of the loss function is:

[0115] ;

[0116] where, represents the value of the loss function, represents the number of samples, represents the th true label of the sample, represents the index of the sample, represents the th predicted value of the sample, represents the logarithmic function;

[0117] The model is tested using the test set, and the model parameters are tuned using the validation set. The optimal parameters are selected according to the performance of the validation set, and early stopping is used to prevent overfitting;

[0118] Through the input layer, hidden layer and output layer architecture of the deep neural network DNN, the compliance assessment model can extract complex non - linear features from a large number of historical compliance assessment datasets, and perform deep learning on the rule matching results, knowledge graph and operation meta - information;

[0119] The deep learning method can simulate complex non - linear relationships, solving the limitation that traditional rule - matching methods cannot fully reflect the potential risks of data operations;

[0120] Compared with traditional linear models and rule - driven methods, DNN can more effectively capture the complex interaction relationships between data features, significantly improving the accuracy and reliability of potential risk assessment, thereby reducing the possibility of misjudgment and missed judgment;

[0121] Through the Softmax activation function, the compliance assessment model can map the potential risk value to the range from 0 to 1, providing an intuitive risk probability value for subsequent risk classification and management of operations;

[0122] By collecting a large number of historical compliance assessment datasets and dividing the datasets into training sets, test sets and validation sets proportionally, the compliance assessment model can be trained and validated on diverse data, ensuring its adaptability to data processing operations in different scenarios;

[0123] The division and training process of the dataset can ensure that the compliance assessment model can not only learn the patterns in historical data, but also effectively respond to newly emerging operation modes and regulatory changes. When a new high - risk operation mode appears, the compliance assessment model can predict risks based on historical similar operation characteristics, avoiding missed judgments of risks caused by changes;

[0124] Evaluate the performance of the compliance assessment model using the test set to further ensure the reliability of the compliance assessment model in practical applications;

[0125] The data-driven training process significantly improves the generalization ability of the compliance assessment model, enabling it to adapt to different data processing scenarios and regulatory rule changes, and ensuring the stable risk prediction ability of compliance assessment in a complex and changing environment;

[0126] Input the data processing features obtained by fusing the rule matching results, knowledge graph, and operation meta-information into the compliance assessment model to calculate the potential risk value. The expression is:

[0127] ;

[0128] Among them, represents the potential risk value, represents the activation function, represents the weight matrix, represents the transpose operation, represents the bias term;

[0129] Set the threshold , when , it means that the current operation is a low-risk operation and no processing is required;

[0130] When , it means that the current operation is a high-risk operation, prevent the current operation from proceeding, and send a warning to the user;

[0131] First, collect a large number of historical compliance assessment data sets, calculate the potential risk value of each historical compliance assessment data set, and set the preliminary threshold by analyzing the distribution of potential risk values;

[0132] According to specific user requirements, adjust the threshold . If the risk cost is high, the threshold can be appropriately reduced to enable the compliance assessment model to give an early warning, which ensures the balance between the sensitivity of the warning and the false alarm rate; the dynamic risk assessment mechanism can classify operations according to the real-time input data features, avoiding the lag in the operation risk assessment in traditional static rules. In high-frequency data processing scenarios, the compliance assessment model can quickly predict risks and take corresponding measures to reduce the threat of high-risk operations to data security and regulatory compliance;

[0133] Through the warning function, it can timely notify the user of potential risks, help the user quickly adjust the operation behavior, and avoid violations and triggering supervision;

[0134] The dynamic risk assessment mechanism significantly improves the real-time nature and emergency response capabilities of compliance assessments. It can not only effectively prevent high-risk operations, but also help users reduce the incidence of illegal operations through accurate early warnings.

[0135] S6. Collect user feedback and optimize the compliance assessment model based on user feedback;

[0136] Calculate data processing feature offset based on user feedback , the expression is:

[0137] ;

[0138] in, Indicates the number of samples of user feedback, The index of the sample representing the user feedback. Indicates The user-annotated feedback value of the sample feedbacked by users, Indicates The potential risk value of the sample reported by users, Indicates Data processing characteristics of samples of user feedback;

[0139] Calculate feature offset based on user feedback Dynamically correct the input features of the compliance assessment model, solving the problem in the existing technology that the compliance assessment model cannot adapt to changes in the distribution of feedback data;

[0140] Add feedback feature offset to the compliance assessment model input layer The correction term redefines the output formula of risk value, and the expression is:

[0141] ;

[0142] represents the adjusted risk value, Indicates feature offset The regulatory factors;

[0143] The scalable compliance assessment model architecture supports the introduction of new data processing features and regulatory rules, and enhances the compliance assessment model's ability to adapt to new scenarios. When regulations are updated and data processing requirements change, the compliance assessment model can quickly adapt to new requirements by adjusting input features.

[0144] The introduction of user feedback further enhances the self-learning ability of the compliance assessment model, ensuring that it maintains efficient risk prediction capabilities in long-term operation;

[0145] The scalability and continuous optimization features of the compliance assessment model enable long-term applicability of compliance assessment. It can be continuously adjusted with changes in regulations and business environments, avoiding problems of aging and performance degradation.

[0146] The revised risk value formula enhances the adaptive ability and assessment accuracy of the compliance assessment model.

[0147] The user feedback optimization mechanism enables the compliance assessment model to continuously learn and improve, ensuring its applicability in the face of regulatory changes and changes in the data processing environment.

[0148] This embodiment also provides a computer device applicable to the case of a data processing method for compliance assessment of data element processing operations, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the data processing method for compliance assessment of data element processing operations as proposed in the above embodiment.

[0149] The computer device can be a terminal. The computer device includes a processor, a memory, a communication interface, a display screen, and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner. The wireless manner can be achieved through WIFI, a carrier network, NFC (Near Field Communication), or other technologies. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer covering the display screen, or a button, trackball, or touchpad set on the outer shell of the computer device, or an external keyboard, touchpad, or mouse, etc.

[0150] This embodiment also provides a storage medium, on which a computer program is stored. When the program is executed by a processor, it implements the data processing method for compliance evaluation for data element processing operations proposed in the above embodiment; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (Static Random Access Memory, abbreviated as SRAM), electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, abbreviated as EEPROM), erasable programmable read-only memory (Erasable Programmable Read Only Memory, abbreviated as EPROM), programmable read-only memory (Programmable Red-Only Memory, abbreviated as PROM), read-only memory (Read-Only Memory, abbreviated as ROM), magnetic memory, flash memory, magnetic disk or optical disc.

[0151] In summary, the present invention solves the problems of inconsistent data and regulatory rule sources and low quality through data query and preprocessing, providing high-quality input data for the entire process. The knowledge graph construction enhances the accuracy and intelligence of rule matching through quantitative correlation and semantic matching. The combination of real-time monitoring and rule matching combines real-time monitoring and dynamic frequency attenuation, improving the accuracy and timeliness of evaluation. The data processing feature extraction generates data features that comprehensively reflect risks through multi-dimensional fusion, providing reliable and efficient input for the compliance evaluation model. Through the key steps of data query preprocessing, knowledge graph construction, real-time monitoring and rule matching, feature extraction, and user feedback optimization, the present invention effectively solves the problems of inaccurate rule matching, lack of dynamic adaptability, and insufficient real-time performance in the prior art, constituting a dynamic, accurate, and efficient compliance evaluation method, which is of great significance to the fields of data processing and compliance management.

[0152] 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 preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

Claims

1. A data processing method for compliance assessment of data element processing operations, characterized in that: include, Collect data elements and regulatory rules and perform pre-processing; Based on the pre-processed data elements and regulations, a knowledge graph is constructed that associates data elements with regulations. Monitor data element processing operations in real time, extract operation meta-information, and perform rule matching based on knowledge graphs; The rule matching results, knowledge graph and operation meta-information are integrated to obtain data processing features; Use machine learning algorithms to build compliance assessment models, input data processing features into the compliance assessment models, analyze potential risks, and issue early warnings and restrictions on high-risk operations; Collect user feedback and optimize the compliance assessment model based on user feedback; The specific steps of constructing the knowledge graph that associates data elements with regulations and rules are as follows: Calculating data elements Regulations and rules specific conditions in The semantic similarity of , the expression is: ； Calculate the correlation between data elements and regulations and rules. The expression is: ； in, Represents data elements, Indicates regulations and rules. Indicates the degree of association, represents the cosine similarity, Indicates regulations and rules The specific conditions in Representing data elements Regulations and rules specific conditions in The semantic similarity of Indicates regulations and rules The weight of Representing data elements The complexity of Indicates regulations and rules complexity; The data element processing operation is monitored in real time, the operation meta-information is extracted, and rule matching is performed in combination with the knowledge graph. The specific steps are as follows: Use the event-driven monitoring tool Kafka to monitor the event stream of data processing operations and capture operation metadata ,in, Indicates the data elements involved in the operation, Indicates the operating frequency; Combined with the knowledge graph for rule matching, the expression is: ； in, Indicates the matching degree, Indicates the data elements involved in the operation and regulations specific conditions in The semantic similarity of represents the natural number base, represents the adjustment coefficient, represents the operating frequency attenuation factor; The rule matching results, knowledge graph and operation meta-information are integrated to obtain data processing features, which are expressed as follows: ； in, Represents data processing characteristics, Indicates the complexity of the data elements involved in the operation, represents the adjustment parameter of the matching degree, represents the adjustment parameter of the operating frequency, Represents the natural logarithm function.

2. The data processing method for compliance assessment of data element processing operations according to claim 1, characterized in that: The data elements and regulations are collected and preprocessed, and the specific steps are as follows: The data elements are user data, device data and transaction data; The regulations and rules refer to the text of laws and regulations related to data elements; Preprocessing of data elements refers to data cleaning, data standardization, and data labeling; Preprocessing of regulations and rules refers to using natural language processing (NLP) to parse and extract text to obtain specific conditions, structured storage, and rule conflict detection.

3. The data processing method for compliance assessment of data element processing operations according to claim 2, characterized in that: The specific steps of using machine learning algorithms to build a compliance assessment model are as follows: Select the deep neural network DNN algorithm to build a compliance assessment model; Setting the architecture of the compliance assessment model based on the deep neural network DNN; The architecture of the compliance assessment model refers to an input layer, a hidden layer, and an output layer; Collecting large historical compliance assessment data sets; The historical compliance assessment data set is divided into training set, test set and validation set in proportion, and input into the compliance assessment model for training, testing and validation of the compliance assessment model respectively.

4. The data processing method for compliance assessment of data element processing operations according to claim 3, characterized in that: The specific steps of inputting data processing characteristics into the compliance assessment model, analyzing potential risks, and issuing early warnings and restrictions on high-risk operations are as follows: The data processing features obtained by integrating rule matching results, knowledge graphs and operation meta-information are input into the compliance assessment model to calculate the potential risk value. The expression is: ； in, represents the potential risk value, express Activation function, represents the weight matrix, represents the transpose operation, represents the bias term; Setting the Threshold ,when When , it indicates that the current operation is a low-risk operation and no processing is performed; when , it indicates that the current operation is a high-risk operation, the current operation is blocked, and an early warning is sent to the user.

5. The data processing method for compliance assessment of data element processing operations according to claim 4, characterized in that: The specific steps of collecting user feedback and optimizing the compliance assessment model based on the user feedback are as follows: Calculate data processing feature offset based on user feedback , the expression is: ； in, Indicates the number of samples of user feedback, The index of the sample representing the user feedback. Indicates The user-annotated feedback value of the sample feedbacked by users, Indicates The potential risk value of the sample reported by users, Indicates Data processing characteristics of samples of user feedback; Add feedback feature offset to the compliance assessment model input layer The correction term redefines the output formula of risk value, and the expression is: ； represents the adjusted risk value, Indicates feature offset The regulating factor.

6. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, it implements the steps of the data processing method for compliance assessment of data element processing operations described in any one of claims 1 to 5.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the data processing method for compliance assessment of data element processing operations described in any one of claims 1 to 5 are implemented.

Images