Data auditing system for terahertz data

Abstract

The invention relates to the technical field of data auditing, and discloses a terahertz data auditing system, comprising a data acquisition module responsible for receiving and preprocessing original data; the data classification module performs automatic classification on the data through a machine learning algorithm, and performs grading according to the sensitivity, the purpose and the source of the data; the permission setting module allocates access permissions to different types of data according to a preset strategy to ensure the security and compliance of the data; the auditing module records and audits all data access operations, and detects an abnormal request; and the access control module is combined with an encryption technology and an identity authentication mechanism, so that the security of data access is ensured, user operation is monitored in real time, and an alarm is triggered to deal with potential security threats. The data acquisition module, the data classification module, the permission setting module, the auditing module and the access control module realize high security and compliance in the data processing and access process of the terahertz data.

Description

Technical Field

[0001] The present invention relates to the technical field of data review, and more particularly, to a data review system for terahertz data. Background Art

[0002] With the continuous development of technology, terahertz data, as an important form of high-precision data, is widely used in fields such as biomedicine, material detection, and environmental monitoring. Due to its characteristics of high resolution and high penetration, terahertz data shows great potential in multiple industries. However, the acquisition and processing of terahertz data face problems such as a large amount of data and high complexity. How to effectively manage and protect these data has become an urgent technical problem to be solved.

[0003] Currently, the management of terahertz data mainly relies on traditional databases and file systems, but this method often has problems such as poor data storage security and lax data access control. Especially in application scenarios with high sensitivity and high confidentiality requirements, terahertz data is vulnerable to unauthorized access and malicious operations. In addition, with the increasing demand for data classification, how to reasonably classify and hierarchically manage these data and ensure the compliance of their access permissions remains a technical challenge.

[0004] Therefore, there is an urgent need to invent a data review technology for terahertz data to solve the problems of poor security and lax access control in the management of terahertz data by traditional databases and file systems. Summary of the Invention

[0005] In view of this, the present invention proposes a data review system for terahertz data, aiming to solve the problems of poor security and lax access control in the management of terahertz data by traditional databases and file systems in the current technology.

[0006] The present invention proposes a data review system for terahertz data, including:

[0007] A data acquisition module, configured to receive and store raw data from terahertz sensors or other data sources, and preprocess the data;

[0008] A data classification module, electrically connected to the data acquisition module, the data classification module is configured to automatically classify terahertz data using machine learning algorithms, and divide the data into different levels according to data sensitivity, usage, and source;

[0009] A permission setting module, electrically connected to the data classification module, the permission setting module is configured to, after the data classification is completed, assign corresponding access permissions to different categories of data according to a preset policy, and set the access permissions of different levels of users;

[0010] The audit module is configured to record all data access and operation behaviors and review abnormal access requests based on the audit policy;

[0011] The access control module is electrically connected to the data classification module, the authority setting module and the audit module respectively. The access control module is configured to ensure data access security based on encryption and identity authentication technology, wherein:

[0012] The access control module is also configured to monitor user operation behavior in real time and trigger a security alarm when abnormal access is detected.

[0013] Furthermore, the data acquisition module pre-processes the data, including:

[0014] The data acquisition module is also configured to remove noise from the raw data based on a filter algorithm and filtering in the time domain or frequency domain;

[0015] The data acquisition module is also configured to convert each original data after noise removal into a preset data format, and remove duplicate data and invalid data from each original data to generate each terahertz data.

[0016] Furthermore, the data classification module uses a machine learning algorithm to automatically classify terahertz data, including:

[0017] The classification module is also configured to extract key features from the terahertz data and establish a classification model based on a decision tree according to the key features;

[0018] The classification module is also configured to train the classification model using supervised learning based on the labeled training data, and adjust the trained classification model through cross-validation and hyperparameter tuning;

[0019] The classification module is further configured to classify each terahertz data based on the adjusted classification model.

[0020] Furthermore, the key features are specifically: the frequency of the terahertz data, the amplitude of the terahertz data and the time characteristics of the terahertz data.

[0021] Furthermore, the classification module builds a classification model based on a decision tree according to key features, including:

[0022] The classification module is also configured to calculate the information gain-based splitting effect of each feature and select the best splitting feature;

[0023] The classification module is also configured to select the best feature for data partitioning based on the calculated information gain value at each decision node;

[0024] The classification module is also configured to recursively divide the data set into subsets at each decision node according to the comparison result of the feature value and the feature threshold until the leaf node;

[0025] The classification module is also configured to finally predict the class of the input data point according to the structure of the decision tree.

[0026] Furthermore, the permission setting module is configured to, after the data classification is completed, assign corresponding access permissions to different categories of data according to a preset policy, including:

[0027] The permission setting module is also configured to define preset user roles and correspond different roles to specific data access permissions;

[0028] The permission setting module is also configured to divide the terahertz data into different access levels according to the data classification result, where the access levels are public data, restricted data, and confidential data;

[0029] The permission setting module is also configured to determine the access permission of the terahertz data according to the access level, data sensitivity, and data source of the terahertz data.

[0030] Furthermore, when the permission setting module is configured to determine the access permission of the terahertz data according to the access level, data sensitivity, and data source of the terahertz data, it includes:

[0031] The permission setting module is also configured to determine the classified score of the terahertz data according to the access level, data sensitivity, and data source of the terahertz data:

[0032] S = A + V + B;

[0033] Where S is the classified score, A is the access level of the terahertz data, V is the data sensitivity of the terahertz data, and B is the data source of the terahertz data;

[0034] The permission setting module is also configured to determine the access permission of the terahertz data according to the classified score.

[0035] Furthermore, when the permission setting module determines the access permission of the terahertz data according to the classified score, it includes:

[0036] The permission setting module is also configured to determine the access permission of the terahertz data according to the relationship between the classified score and the configured first preset classified score and second preset classified score:

[0037] When the classified score is lower than or equal to the first preset classified score, the permission setting module determines that the access permission of the terahertz data is at a low level;

[0038] When the classified score is higher than the first preset classified score and lower than or equal to the second preset classified score, the permission setting module determines that the access permission for the terahertz data is medium level;

[0039] When the classified score is higher than the second preset classified score, the permission setting module determines that the access permission for the terahertz data is high level;

[0040] Among them, the first preset classified score is lower than the second preset classified score, and the access permissions are, from high to low, high level, medium level, and low level.

[0041] Further, when the audit module reviews the abnormal access request based on the audit policy, it includes:

[0042] The audit module is also configured to identify and mark the access requests that deviate from the normal mode as objections by setting the normal access mode, and the audit module is also configured to conduct a secondary review of the objection access requests;

[0043] The audit module is also configured to obtain the historical access behavior of the user of the objection access request, and determine whether the access request of the user of the objection access request is an abnormal access request according to the historical access behavior.

[0044] Further, when the audit module determines whether the access request of the user of the objection access request is an abnormal access request according to the historical access behavior, it includes:

[0045] The audit module is also configured to determine the normal access mode of the user of the objection access request according to the historical access behaviors of the user of the objection access request, and compare the difference between the normal access mode and the access behavior of the abnormal access request;

[0046] If the normal access mode is consistent with the access behavior of the abnormal access request, the audit module determines that the objection access request is not an abnormal access request;

[0047] If the normal access mode is inconsistent with the access behavior of the abnormal access request, the audit module determines that the objection access request is an abnormal access request and conducts a secondary identity verification on the user of the objection access request;

[0048] Among them, the secondary identity verification includes SMS verification code, email confirmation, and biometric verification, and the access behavior includes login information, access frequency, access time, and sensitivity of the accessed data.

[0049] Compared with the prior art, the beneficial effects of the present invention are as follows: By integrating modules such as data acquisition, classification, permission setting, auditing, and access control, the security and management efficiency of terahertz data can be effectively improved. The data acquisition module ensures the accurate storage and preprocessing of raw data. The data classification module realizes the intelligent classification of data through machine learning algorithms, providing a basis for subsequent permission allocation and security control. The permission setting module configures reasonable access permissions according to the sensitivity, usage, and source of the data, ensuring that different users can only access data within their permission scope. The auditing module can record and review all data access and operation behaviors, promptly detect abnormal access, and avoid potential security threats. At the same time, the access control module uses encryption and identity authentication technologies to ensure the security of data access, and through real-time monitoring and alert mechanisms, automatically responds in case of abnormal access, effectively preventing data leakage and improper operations, and ensuring the compliance and security of data management. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] By reading the following detailed description of the preferred embodiments, various other advantages and benefits will become clear to those of ordinary skill in the art. The drawings are only for the purpose of showing the preferred embodiments and are not to be considered as limiting the present invention. Moreover, throughout the drawings, the same reference numerals are used to represent the same components. In the drawings:

[0051] Figure 1 is a functional block diagram of a data auditing system for terahertz data provided by an embodiment of the present invention; DETAILED DESCRIPTION OF THE EMBODIMENTS

[0052] Hereinafter, the exemplary embodiments of the present disclosure will be described in more detail with reference to the drawings. Although the exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided so that the present disclosure can be more thoroughly understood and the scope of the present disclosure can be completely conveyed to those skilled in the art. It should be noted that, without conflict, the embodiments and features in the embodiments of the present invention can be combined with each other. The present invention will be described in detail below with reference to the drawings and in conjunction with the embodiments.

[0053] As Figure 1 shown, in some embodiments of the present application, this embodiment provides a data auditing system for terahertz data, including: a data acquisition module, a data classification module, a permission setting module, an auditing module, and an access control module.

[0054] Specifically, the data acquisition module is configured to receive and store the raw data from terahertz sensors or other data sources and preprocess the data; the data classification module is electrically connected to the data acquisition module, and the data classification module is configured to automatically classify the terahertz data using machine learning algorithms and divide the data into different levels according to data sensitivity, usage, and source; the permission setting module is electrically connected to the data classification module, and the permission setting module is configured to, after the data classification is completed, allocate corresponding access permissions to different categories of data according to preset policies and set the access permissions for different levels of users; the audit module is configured to record all data access and operation behaviors and review abnormal access requests based on audit policies; the access control module is electrically connected to the data classification module, the permission setting module, and the audit module respectively, and the access control module is configured to ensure data access security based on encryption and identity authentication technologies. Among them, the access control module is further configured to monitor user operation behaviors in real time and trigger a security alarm when detecting abnormal access.

[0055] It can be understood that the data acquisition module is responsible for receiving and storing the raw data from terahertz sensors or other data sources and preprocessing this data. The preprocessing includes steps such as denoising, redundant data removal, and data format conversion to ensure that the collected raw data has high quality before transmission and storage, providing a reliable basis for subsequent processing and analysis. Secondly, the preprocessed terahertz data is automatically classified by using machine learning algorithms. By analyzing the sensitivity, usage, and source of the data, the data is divided into different levels. This provides classification information for the permission setting module, which allocates corresponding access permissions to different levels of data according to preset policies and sets the access permissions for different levels of users at the same time to ensure the compliance and security of data access. Finally, the access control module combines encryption and identity authentication technologies to ensure the security of data access. This module can monitor user operation behaviors in real time, detect and respond to abnormal access behaviors in a timely manner, and trigger a security alarm. The audit module is responsible for recording all data access and operation behaviors, reviewing abnormal access according to audit policies, and further enhancing the security and transparency of data management.

[0056] It can be seen that the data acquisition module removes noise and redundant data through preprocessing and standardizes the data format, ensuring the accuracy and consistency of subsequent processing and analysis. This can effectively improve the quality of terahertz data, avoid misjudgment or erroneous processing due to data problems, and provide a reliable data basis for subsequent classification and permission setting. Secondly, the data classification module realizes the automatic classification of terahertz data through machine learning algorithms, enabling the system to intelligently identify and process different types of data. The permission setting module accurately controls the access rights of users at different levels based on the classification results and preset strategies, ensuring that the sensitivity and use of data are reasonably protected and avoiding improper access or leakage. Finally, the combination of the access control module and the audit module not only makes data access subject to strict permission control, but also monitors the user's operation behavior in real time. By triggering security alarms and reviewing abnormal access, the system can quickly respond to potential security threats. This mechanism significantly improves the security of data management and ensures that the system can automatically respond to and prevent non-compliant or malicious access behaviors.

[0057] Specifically, when the data acquisition module preprocesses the data, it includes: the data acquisition module is also configured to remove noise from the original data based on a filter algorithm and time domain or frequency domain filtering; the data acquisition module is also configured to convert the noise-removed original data into a preset data format, and eliminate duplicate data and invalid data in each original data to generate each terahertz data.

[0058] It can be seen that the data acquisition module effectively removes noise from the original data based on filter algorithms and time domain or frequency domain filtering technology, thereby improving the signal-to-noise ratio and quality of the data. At the same time, by converting the denoised data into a preset standard format and eliminating redundant data and invalid data, the generated terahertz data is ensured to be consistent and accurate. This process not only improves the availability and reliability of the data, but also provides a high-quality data foundation for subsequent data classification, permission setting and review, thereby enhancing the data management efficiency and security of the entire system.

[0059] Specifically, when the data classification module uses a machine learning algorithm to automatically classify terahertz data, it includes: the classification module is also configured to extract key features from the terahertz data, and establish a classification model based on the key features and a decision tree; the classification module is also configured to train the classification model based on the labeled training data using a supervised learning method, and adjust the trained classification model through cross-validation and hyperparameter tuning; the classification module is also configured to classify each terahertz data based on the adjusted classification model.

[0060] Specifically, the key features are: the frequency of the terahertz data, the amplitude of the terahertz data, and the temporal characteristics of the terahertz data.

[0061] Specifically, when the classification module builds a classification model based on a decision tree according to key features, it includes: The classification module is also configured to calculate the splitting effect of each feature based on information gain and select the best splitting feature; The classification module is also configured to select the best feature for data partitioning at each decision node according to the calculated information gain value; The classification module is also configured to recursively divide the data set into subsets at each decision node according to the comparison result of the feature value and the feature threshold until the leaf node; The classification module is also configured to finally predict the class of the input data point according to the structure of the decision tree.

[0062] It can be understood that

[0063] It can be seen that

[0064] Specifically, when the permission setting module is configured to assign corresponding access permissions to different categories of data according to a preset policy after data classification, it includes: The permission setting module is also configured to define preset user roles and correspond different roles to specific data access permissions; The permission setting module is also configured to divide terahertz data into different access levels according to the data classification result, where the access levels are public data, restricted data, and confidential data; The permission setting module is also configured to determine the access permissions of terahertz data according to the access level, data sensitivity, and data source of the terahertz data.

[0065] Specifically, when the permission setting module is also configured to determine the access permissions of terahertz data according to the access level, data sensitivity, and data source of the terahertz data, it includes: The permission setting module is also configured to determine the classified score of the terahertz data according to the access level, data sensitivity, and data source of the terahertz data: S = A + V + B; where S is the classified score, A is the access level of the terahertz data, V is the data sensitivity of the terahertz data, and B is the data source of the terahertz data; The permission setting module is also configured to determine the access permissions of the terahertz data according to the classified score.

[0066] Specifically, when the permission setting module determines the access permission of terahertz data according to the classified information score, it includes: the permission setting module is further configured to determine the access permission of terahertz data according to the relationship between the classified information score and the configured first preset classified information score and the second preset classified information score: when the classified information score is lower than or equal to the first preset classified information score, the permission setting module determines that the access permission of terahertz data is a low level; when the classified information score is higher than the first preset classified information score and lower than or equal to the second preset classified information score, the permission setting module determines that the access permission of terahertz data is a medium level; when the classified information score is higher than the second preset classified information score, the permission setting module determines that the access permission of terahertz data is a high level; wherein, the first preset classified information score is lower than the second preset classified information score, and the access permissions are, from high to low, high level, medium level and low level.

[0067] It can be understood that the data classification module first extracts key features from the terahertz data, including frequency, amplitude, and time characteristics, which are of great significance for data classification. Subsequently, based on the extracted key features, the classification module uses the decision tree algorithm to establish a classification model. The decision tree algorithm selects the best feature for splitting by calculating the information gain of each feature, and recursively divides the data set into multiple subsets, finally generating a classification tree structure. Secondly, through the labeled training data, the decision tree classification model is trained in a supervised learning manner. The generalization ability of the model is evaluated through cross-validation techniques, and the parameters of the decision tree are adjusted through hyperparameter tuning methods to optimize the classification effect of the model, avoid overfitting or underfitting problems, and thus improve the classification accuracy. Finally, according to the trained decision tree model, the terahertz data is classified. When a data point passes through the decision tree, each decision node selects the most appropriate feature for data division according to the calculated information gain value, and finally gives the predicted class of the data point at the leaf node. This process ensures the accurate classification of the data through recursive division, thereby effectively dividing different types of terahertz data into predetermined classes.

[0068] It can be seen that by extracting key features (frequency, amplitude, time characteristics) from terahertz data, the data classification module can make full use of the internal laws of the data to improve the accuracy and efficiency of classification. Feature selection can ensure that the classification model better captures the main information of the data and optimizes the classification effect. Secondly, by adopting the method of supervised learning and techniques such as cross-validation and hyperparameter tuning, the classification model can be accurately trained with the support of labeled data, avoiding overfitting or underfitting of the model. In this way, the classification model can adapt to different terahertz data characteristics and has strong generalization ability, with higher accuracy in practical applications. Finally, the decision tree algorithm optimally selects features and partitions data through information gain, making the classification process highly efficient and accurate. The way of recursively partitioning the data set enables each data point to be accurately classified through the decision tree, ensuring that different categories of terahertz data can be effectively identified and processed, thus providing reliable support for subsequent permission setting and data management.

[0069] Specifically, when the audit module reviews abnormal access requests based on the audit policy, it includes: the audit module is also configured to identify and mark access requests that deviate from the normal mode as objections by setting a normal access mode, and the audit module is also configured to conduct a secondary review of the objection access requests; the audit module is also configured to obtain the historical access behaviors of the users of the objection access requests and determine whether the access requests of the users of the objection access requests are abnormal access requests based on the historical access behaviors.

[0070] Specifically, when the audit module determines whether the access request of the user of the objection access request is an abnormal access request based on the historical access behaviors, it includes: the audit module is also configured to determine the normal access mode of the user of the objection access request based on the historical access behaviors of the user of the objection access request and compare it with the access behaviors of the abnormal access request; if the normal access mode is consistent with the access behaviors of the abnormal access request, the audit module determines that the objection access request is not an abnormal access request; if the normal access mode is inconsistent with the access behaviors of the abnormal access request, the audit module determines that the objection access request is an abnormal access request and conducts secondary authentication on the user of the objection access request; among them, the secondary authentication includes SMS verification codes, email confirmations, and biometric verifications, and the access behaviors include login information, access frequency, access time, and the sensitivity of the accessed data.

[0071] It is understandable that the access behavior standard of users is established by setting a normal access mode. The normal mode includes characteristics such as the user's login time, access frequency, and sensitivity of the accessed data. When the user's access request deviates from this normal mode, the system will mark it as a "disputed" access request and initiate a further review process. This process relies on the learning and modeling of the user's historical access behavior to provide a basis for judging deviant behavior. Secondly, the audit module obtains the historical access behavior of the user with the disputed access request and determines the user's normal access mode based on this historical data. By comparing the differences between the normal mode and the disputed access request, the audit module can judge whether the request is an abnormal access. If the difference is large, it indicates that the access request may be abnormal, and the audit module will further perform secondary authentication to confirm the legitimacy of the access request. Finally, when the audit module detects a disputed access request, it conducts further identity verification through secondary authentication. Secondary authentication may include means such as SMS verification codes, email confirmations, and biometric verification. In addition, key factors of access behavior (such as login information, access frequency, access time, and sensitivity of accessed data) will be used for analysis to ensure that only compliant and verified access can continue. This mechanism can effectively prevent unauthorized access and malicious behavior.

[0072] It can be seen that by setting a normal access mode and identifying access requests that deviate from the normal mode, the system can automatically discover potential security risks and abnormal behaviors. Only when the access request conforms to the normal mode can it be regarded as a legitimate request, thus effectively avoiding malicious attacks or unauthorized access behaviors and enhancing the overall security of the system. Secondly, by comparing the disputed access request with the historical access behavior, the audit module can more accurately judge which requests are abnormal. This comparative analysis based on historical behavior can improve the detection rate of abnormal access, avoid misjudgment caused by a single abnormal event, and ensure the compliance and legitimacy of data access. Finally, when an abnormal access request is detected, multiple authentication means (such as SMS verification codes, email confirmations, and biometric verification) are used to ensure the authenticity of the visitor's identity. This secondary verification mechanism effectively prevents identity forgery and unauthorized access. Especially in scenarios of accessing highly sensitive data, it can further improve security and ensure that only authorized users can access and operate.

[0073] In the above embodiments, by combining modules such as data collection, classification, permission setting, auditing, and access control, the security and management efficiency of terahertz data can be effectively improved. The data collection module ensures the accurate storage and preprocessing of raw data. The data classification module realizes the intelligent classification of data through machine learning algorithms, providing a basis for subsequent permission allocation and security control. The permission setting module configures reasonable access permissions according to the sensitivity, usage, and source of the data, ensuring that different users can only access data within their permission scope. The auditing module can record and review all data access and operation behaviors, promptly detect abnormal access, and avoid potential security threats. At the same time, the access control module uses encryption and identity authentication technologies to ensure the security of data access. Through real-time monitoring and alarm mechanisms, it automatically responds in case of abnormal access, effectively preventing data leakage and improper operations, and ensuring the compliance and security of data management.

[0074] Those skilled in the art should understand that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0075] The present application is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or block in the flowcharts and / or block diagrams can be implemented by computer program instructions, and the combination of processes and / or blocks in the flowcharts and / or block diagrams can also be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for implementing the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0076] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured article including an instruction device, and the instruction device implements the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0077] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus, so that a series of operation steps are executed on the computer or other programmable apparatus to produce a computer-implemented process, thereby providing instructions for implementing the steps of the process Figure 1 one process or a plurality of processes and / or blocks Figure 1 steps of the functions specified in one block or a plurality of blocks.

[0078] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: the specific embodiments of the present invention can still be modified or equivalently replaced, and any modification or equivalent replacement without departing from the spirit and scope of the present invention shall be covered by the protection scope of the claims of the present invention.

Claims

1. A data audit system for terahertz data, characterized in that, include: A data acquisition module is configured to receive and store raw data from a terahertz sensor or other data source, and pre-process the data; A data classification module, electrically connected to the data acquisition module, the data classification module is configured to automatically classify the terahertz data using a machine learning algorithm, and divide the data into different levels according to data sensitivity, usage and source; The permission setting module is electrically connected to the data classification module. The permission setting module is configured to assign corresponding access rights to different categories of data according to a preset strategy after the data classification is completed, and to set access rights for users of different levels; The audit module is configured to record all data access and operation behaviors and review abnormal access requests based on the audit policy; The access control module is electrically connected to the data classification module, the authority setting module and the audit module respectively. The access control module is configured to ensure data access security based on encryption and identity authentication technology, wherein: The access control module is also configured to monitor user operation behavior in real time and trigger a security alarm when abnormal access is detected.

2. The data review system for terahertz data according to claim 1, characterized in that, The data acquisition module pre-processes the data, including: The data acquisition module is also configured to remove noise from the raw data based on a filter algorithm and filtering in the time domain or frequency domain; The data acquisition module is also configured to convert each original data after noise removal into a preset data format, and remove duplicate data and invalid data from each original data to generate each terahertz data.

3. The data review system for terahertz data according to claim 2, characterized in that, The data classification module uses machine learning algorithms to automatically classify terahertz data, including: The classification module is also configured to extract key features from the terahertz data and establish a classification model based on a decision tree according to the key features; The classification module is also configured to train the classification model using supervised learning based on the labeled training data, and adjust the trained classification model through cross-validation and hyperparameter tuning; The classification module is further configured to classify each terahertz data based on the adjusted classification model.

4. The data review system for terahertz data according to claim 3, wherein, The key features are: the frequency of terahertz data, the amplitude of terahertz data and the time characteristics of terahertz data.

5. The data review system for terahertz data according to claim 4, wherein The classification module builds a classification model based on a decision tree according to key features, including: The classification module is also configured to calculate the information gain-based splitting effect of each feature and select the best splitting feature; The classification module is also configured to select the best feature for data partitioning based on the calculated information gain value at each decision node; The classification module is further configured to recursively divide the data set into subsets at each decision node according to the comparison result between the feature value and the feature threshold until a leaf node; The classification module is also configured to ultimately predict categories for input data points based on the structure of the decision tree.

6. The data review system for terahertz data according to claim 1, wherein The permission setting module is configured to assign corresponding access rights to different categories of data according to preset policies after data classification is completed, including: The permission setting module is also configured to define preset user roles, mapping different roles to specific data access permissions; The permission setting module is further configured to divide the terahertz data into different access levels according to the data classification result, where the access levels are public data, restricted data, and confidential data; The permission setting module is further configured to determine the access permission of the terahertz data according to the access level, data sensitivity, and data source of the terahertz data.

7. The data review system for terahertz data according to claim 6, wherein When the permission setting module is further configured to determine the access permission of the terahertz data according to the access level, data sensitivity, and data source of the terahertz data, it includes: The permission setting module is further configured to determine the classified rating of the terahertz data according to the access level, data sensitivity, and data source of the terahertz data: S = A + V + B; where S is the classified rating, A is the access level of the terahertz data, V is the data sensitivity of the terahertz data, and B is the data source of the terahertz data; The permission setting module is further configured to determine the access permission of the terahertz data according to the classified rating.

8. The data review system for terahertz data according to claim 7, wherein When the permission setting module determines the access permission of the terahertz data according to the classified rating, it includes: The permission setting module is further configured to determine the access permission of the terahertz data according to the relationship between the classified rating and the configured first preset classified rating and second preset classified rating: When the classified rating is lower than or equal to the first preset classified rating, the permission setting module determines that the access permission of the terahertz data is a low level; When the classified rating is higher than the first preset classified rating and lower than or equal to the second preset classified rating, the permission setting module determines that the access permission of the terahertz data is a medium level; When the classified rating is higher than the second preset classified rating, the permission setting module determines that the access permission of the terahertz data is a high level; where the first preset classified rating is lower than the second preset classified rating, and the access permissions are, from high to low, high level, medium level, and low level.

9. The data review system for terahertz data according to claim 1, wherein, When the audit module reviews the abnormal access request based on the audit policy, it includes: The audit module is further configured to identify and mark the access requests that deviate from the normal mode as objections by setting the normal access mode, and the audit module is further configured to conduct a secondary review of the objection access requests; The audit module is further configured to obtain the historical access behavior of the user of the objection access request and determine whether the access request of the user of the objection access request is an abnormal access request according to the historical access behavior.

10. The data review system for terahertz data according to claim 9, characterized in that, When the audit module determines whether the access request of the user of the objection access request is an abnormal access request according to the historical access behavior, it includes: The audit module is further configured to determine the normal access mode of the user of the objection access request according to each historical access behavior of the user of the objection access request, and compare the difference between the normal access mode and the access behavior of the abnormal access request; If the normal access mode is consistent with the access behavior of the abnormal access request, the audit module determines that the objection access request is not an abnormal access request; If the normal access mode is inconsistent with the access behavior of the abnormal access request, the audit module determines that the objection access request is an abnormal access request and conducts a secondary identity verification on the user of the objection access request; Among them, two-factor authentication includes SMS verification codes, email confirmations, and biometric verification, and access behaviors include login information, access frequency, access time, and the sensitivity of the accessed data.

Images