System and method for visible and safe use of data

Abstract

The invention discloses a system and a method for visible and safe use of data. The system comprises user two-way input equipment, an input information storage server, a service operation module and a safety detection module, a service operation server / container is constructed, a user inputs information through a two-way input device, the input information is respectively transmitted to a service operation module and an input information storage server, and a safety detection module respectively extracts and checks user use operation behaviors in the service operation module and the input information in the input information storage server. And when an exception is found, corresponding response operation is carried out according to a security policy. According to the invention, safe use of the authorization data can be ensured, and safe use of the authorization data in an'available and visible 'mode is supported.

Description

Technical Field

[0001] The present invention relates to a method for the safe use of network data, and in particular to a system and method for the visible and safe use of data. Background Art

[0002] With the widespread use of the Internet, how to achieve effective data value integration in a "available but invisible" way while ensuring the data privacy and security of all parties has become a key issue that the data industry urgently needs to solve.

[0003] Traditional data circulation and use solutions mainly involve obtaining and using data through data authorization downloads, API calls, or file exchanges. This method has almost no restrictions on the use of data. After obtaining the data, the data user can use it beyond the permitted scope, copy, tamper with, illegally disseminate, or even resell the original data, causing huge economic and reputational losses to the data owner. The main disadvantages of traditional data circulation and use solutions are as follows: 1. Most systems are based only on username and password authentication systems. The authentication and identification measures are too simple and cannot prevent identity theft caused by account and password leaks.

[0004] 2. Once the data is acquired, it is out of the control of the data provider, and the data user can copy it at will and spread it to other areas.

[0005] 3. It is impossible to monitor the reasonable use of data, and the data user may perform operations beyond authorization.

[0006] 4. When illegal use of data is discovered, there is a lack of technical means to terminate the use of the data.

[0007] 5. After a data security incident occurs, there is a lack of auditing methods to determine the point of the accident.

[0008] In order to address the above technical shortcomings, privacy computing-related technologies have been developed. Privacy computing is a cross-disciplinary technology that covers many disciplines. Mainstream privacy computing technologies are classified into three major directions: the first category is privacy computing technology based on cryptography, represented by multi-party secure computing; the second category is technologies derived from the integration of artificial intelligence and privacy protection technologies, represented by federated learning; the third category is privacy computing technology based on trusted hardware, represented by trusted execution environments. Different technologies are often used in combination to complete data computing and analysis tasks while ensuring the security and privacy of original data.

[0009] At present, mainstream privacy computing technologies include the following methods: Secure Multi-party Computation (MPC) was created by Yao Qizhi in 1982 by proposing and solving the millionaire's problem. It means that in the absence of a trusted third party, multiple parties jointly calculate an objective function and ensure that each party only obtains its own calculation results. It is impossible to infer the input data of any other party through the interactive data in the calculation process (unless the function itself can infer the input of other participants from its own input and the output obtained).

[0010] Federated Learning (FL), also known as federated machine learning, joint learning, and alliance learning, is a multi-party joint machine learning training that is achieved by circulating and processing intermediate encrypted data without leaving the local original data warehouse. Federated learning participants generally include data parties, algorithm parties, coordination parties, calculation parties, result parties, and task initiators. According to the different distribution of data participating in the calculation among data parties, it can be divided into horizontal federated learning, vertical federated learning, and federated transfer learning.

[0011] Trusted Execution Environment (TEE) builds a secure area in the CPU through software and hardware methods to ensure that the programs and data loaded inside are protected in terms of confidentiality and integrity. TEE is an isolated execution environment that provides a higher level of security than the ordinary operating system (Rich Operating System, RichOS) and more functions than the secure element (Secure Element, SE) for trusted applications running on the device.

[0012] Multi-party intermediary computation (MPIC) is a new privacy computing method proposed by Tan Li and Kong Jun. It refers to a data processing method in which multi-party data is analyzed and calculated and anonymized results are output through a secure and reliable mechanism in a regulated intermediary computing environment independent of the data party and the user. It is a computing management system. In MPIC, the original data of the data party is de-identified and input into the intermediary computing environment or platform for computing. It is deleted immediately after the computing is completed. The anonymized result data is output according to the specified path after review. Under the specific environment and rules of MPIC, the identity of the information data is encrypted and identified. Because its algorithm is irreversible, it cannot be restored to the original data, which meets one requirement of anonymization, that is, it cannot be restored; at the same time, because these de-identified information data are closed in a specific regulated environment or platform, it objectively meets another requirement of anonymization, that is, it is impossible to identify a specific natural person. Therefore, the processed data can be regarded as anonymized in essence and no longer belongs to personal information. It can enter the intermediary computing environment or platform for computing without obtaining personal consent.

[0013] Although the above mainstream privacy computing can solve the problems in traditional data circulation to a certain extent, they also have their own shortcomings: 1. Multi-party secure computing uses a large number of cryptographic algorithms, which leads to bottlenecks in computing performance and is several times or even dozens of times slower than plaintext computing.

[0014] 2. Federated computing has the possibility of inferring the data distribution of participants, and it is difficult to avoid the possibility of malicious participants providing false data. In addition, the overall progress may be affected due to inconsistent computing capabilities of various participants.

[0015] 3. The trusted execution environment requires a high hardware investment and is closely tied to international CPU manufacturers, which will affect the credibility of the trusted execution environment technology.

[0016] 4. Multi-party intermediary computing is input into the intermediary computing environment or platform to participate in computing after de-identification, making it impossible to implement applications that require the use of identity associations.

[0017] At the same time, the above mainstream privacy computing technologies also have the following defects: 1. Unable to prove safety by oneself, it is necessary to find an authoritative organization to endorse and certify. Safety assessment or qualification certification requires a large cost investment.

[0018] 2. It cannot solve the authorization problem. The essence of privacy computing is to solve the problem of secure data sharing among multiple parties. It does not have the authorization function itself and needs to be combined with other technologies such as blockchain to solve the authorization problem.

[0019] 3. Privacy computing ensures the secure sharing of data among all parties through a "available but invisible" approach, but is powerless in the scenario where "available, visible and secure" is required.

[0020] Therefore, it is necessary to design a method that ensures data privacy and security while enabling the safe use of authorized data in a "available and visible" manner. Summary of the invention

[0021] In order to solve the above problems, the present invention proposes a system and method for data available, visible and safe use, which can ensure the safe use of authorized data and support the safe use of authorized data in a "available and visible" manner.

[0022] The technical solution of the present invention is as follows: A system for data to be visible and used safely includes a user dual-channel input device, an input information storage server, a business operation module, and a security detection module.

[0023] The user dual-channel input device includes at least one input device such as a user local terminal, a keyboard, a mouse, a screen touch device, an input splitter, etc., which is used to output the user's operation in two channels, one channel connected to the business operation server / container, and the other channel connected to the input information storage server.

[0024] The input information storage server is used to store the unique identification number of the input device, input operation information (such as keyboard / mouse keystroke information, touch screen / touch input device click information) and time information for subsequent security detection.

[0025] The business operation module includes a business operation server / container and business software for using data; the business operation server / container is used to provide a closed data use environment to ensure that the data loaded in the closed data use environment is closed and used safely. Specifically, the startup of the business operation server can be built in advance for standby or built according to user needs, and the business operation container can be pulled up according to user needs.

[0026] Furthermore, when the business operation server / container is working: When the business operation container is pulled up or the business operation server is started, the business software and the encrypted data to be used are loaded, and the data to be used is decrypted after obtaining the authorization key. After obtaining authorization, the data is loaded into the data closed use environment. The user can use the data by remotely accessing the data closed use environment. During the entire use process, the data is prohibited from leaving the data closed use environment. Each time the business operation server / container works, a corresponding business operation log will be generated.

[0027] In special circumstances (such as when the parties involved are highly trusted), the data authorization and decryption process can be omitted. After the business operation container / server is started, the user can use the data by remotely accessing the closed data environment.

[0028] The data closed use environment can be achieved through a dedicated hardware server with security measures; the dedicated hardware server refers to a server environment established for the data usage scenario, using necessary security network configuration combined with security protection software to achieve an environment where data cannot be arbitrarily transferred outside the server, such as through a corresponding network disconnection mechanism to achieve data closure. It is also possible to use virtual machine technology with security measures to build a secure container to achieve a data closed use environment, or use containerization technology such as Docker with security measures to build a secure container to achieve a data closed use environment.

[0029] Furthermore, the business software selects mature software or customized special software according to the scenario of data use, and creates value through data mining.

[0030] The security detection module, when running the security detection program, is used to check whether the user's behavior in the business operation module is normal. When an abnormality is found, the corresponding response operation is performed according to the security policy configuration, and the user operation can be interrupted immediately, and the data can be destroyed if necessary (for example, when there is still a risk of data leakage after the interruption of the operation, the data destruction measure is adopted). It can also prompt abnormalities, request three-party confirmation, send warning information and other different operation modes. Among them: the behavior of the inspection includes but is not limited to whether the user account and the dual-channel input device identifier match, whether the user performs illegal operations, whether the user operation and input information match, whether the user's keystroke behavior characteristics match the person, etc.; the inspection strategy is selected or customized according to business needs, and can adopt fixed-point inspection, or regular inspection, or fixed-point and regular inspection, or continuous inspection, etc.

[0031] Based on the above system, a method for achieving fully controlled and secure use of authorized data includes the following steps: Step 1: Build a business operation server / container to provide a closed data usage environment. Data users obtain authorization and load authorized data into the closed data usage environment.

[0032] Step 2: The input information generated by the user through the dual-way input device is connected to the business operation server / container through the first channel to perform data usage operations; when the user performs business operations, the input information generated by the dual-way input device is transmitted to the input information storage server through the second channel for storage; the stored information includes the unique identifier of the input device, the input content, the input time, etc.; the original input information transmitted through the first channel and the second channel is the same.

[0033] Step 3: The security detection module extracts the user's usage operation behavior in the business operation module and the input information in the input information storage server, and checks the extracted usage operation behavior and input information according to the set inspection strategy. If an abnormality is found, a response operation is taken according to the configuration.

[0034] Further restrictions on the steps for the above authorized data to be used in a fully controlled and secure manner: In step 3, the use operation behavior in the business operation module is extracted by monitoring various system logs or parsing the received business operation data packets.

[0035] In step 3, the extracted usage operation behavior and input information are continuously checked as follows: (1) Check whether the user account and the unique identifier of the input device match.

[0036] (2) Check whether the user's operation behavior contains any illegal operations such as illegal copying, external transmission, etc.

[0037] (3) Check whether the user's operation behavior matches the input information, such as keyboard and mouse keystroke information.

[0038] (4) AI technology can be combined to check whether the user's operating behavior characteristics match the user's own.

[0039] (5) Conduct risk assessment on the above inspection results, and the assessment results may include: interrupting user operations and destroying data.

[0040] In the above steps, the security detection module is also used to archive information such as user operation behavior, keystroke information and inspection process records to provide data support for auditing and backtracking; the information stored in the input information storage server, combined with the business operation log, can be used to review illegal operations in the data usage process and trace the user's violation responsibility.

[0041] The beneficial effects of the present invention are as follows: 1. The present invention has a wide range of applications, high versatility, and low cost.

[0042] Second, the present invention adopts a continuous two-factor authentication implementation scheme, which can greatly improve the security of the system, especially the data usage process, and can detect illegal behaviors on demand.

[0043] 3. The present invention places the use of data in a safe and closed environment, monitors the identity and operation behavior of users in real time, and can promptly interrupt user operations or even destroy data when security risks are found. The data operation behavior, keyboard and mouse keystroke information, security inspection process and other information will be retained for a certain period of time, which can provide data support during auditing and backtracking. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 It is a system schematic diagram of the present invention. DETAILED DESCRIPTION

[0045] The present invention designs a system for data to be visible and safe to use, including a user dual-channel input device, an input information storage server, a business operation module, and a security detection module. Among them: The user dual-path input device is used to output the user's operation in two paths, one path connected to the business operation server / container, and the other path connected to the input information storage server.

[0046] The input information storage server is used to store the unique identification number of the input device, input operation information and time and other information for subsequent security detection.

[0047] The business operation module includes a business operation server / container and business software for using data; the business operation server / container is used to provide a closed data use environment to ensure that the data loaded in the closed data use environment is used closed and safely.

[0048] The security detection module is used to arrange fixed-point, regular or continuous checks on whether the user's behavior in the business operation module is normal during the security detection process according to business needs; when an abnormality is found, the corresponding response operation is performed according to the security policy. Among them, the checked behavior includes one or more of whether the user account and the dual-channel input device identifier match, whether the user has performed illegal operations, whether the user operation and input information match, and whether the user's keystroke behavior characteristics match the person himself. The security detection module is also used to archive information such as user operation behavior, keystroke information, and inspection process records to provide data support for auditing and backtracking; the information stored in the input information storage server, combined with the business operation log, can be used to review illegal operations in the data usage process and trace the user's violation responsibility.

[0049] Traditional user local terminals are ordinary PCs or dedicated terminal boxes, etc., which are used as remote access clients to connect to the remote access service provided by the business operation container. The user input can be divided into two outputs using software or hardware. Remote access methods include remote desktop (windows) or SSH (linux), etc. However, if the terminal is attacked and taken down by hackers, the security measures of dual-channel monitoring are likely to be cracked.

[0050] Therefore, in this solution, for dual-channel hardware input devices such as dual-channel keyboards and dual-channel mice, the unique identification number of the input device is bound to the user account. In addition to normal channel connection (such as keyboard, mouse and other input devices, connected to the user's local terminal, and then through the terminal's network card and other network devices) for business operations, another set of channels is used to transmit the unique identification number of the input device, input operation information (such as keyboard, mouse keystroke information), etc. to the input information storage server for storage, so that it can be obtained and used by the security detection container later.

[0051] Based on the above system, a method for achieving fully controlled and secure use of authorized data includes the following steps: Step 1: Build a business operation server / container to provide a closed data usage environment. Data users obtain authorization and load authorized data into the closed data usage environment.

[0052] Step 2: The input information generated by the user through the dual-way input device is connected to the business operation server / container through the first channel to perform data usage operations; when the user performs business operations, the input information generated by the dual-way input device is transmitted to the input information storage server through the second channel for storage; the stored information includes the unique identifier of the input device, the input content, the input time, etc.; the original input information transmitted through the first channel and the second channel is the same.

[0053] Step 3: The security detection module extracts the user's usage operation behavior in the business operation module and the input information in the input information storage server, and checks the extracted usage operation behavior and input information according to the set inspection strategy. If an abnormality is found, a response operation is taken according to the configuration.

[0054] According to the above system and method, the specific implementation scheme may be as follows: 1. How to implement the container 1.1 Remote access to container The user logs in to the business platform using a dedicated keyboard and mouse, and the login account, password, and verification code are sent to the business system for login verification; the keyboard unique identifier and keystroke information are sent to the keyboard and mouse signal storage server, and the system security detection module checks whether the verification code matches the keystroke information of the user's dedicated keyboard, whether the user account and the keystroke keyboard unique identifier match, and whether the keystroke information matches the user's behavioral characteristics. The login is successful after all the above checks are passed.

[0055] After the user logs in, the business platform container is opened on the business platform scenario page: when it is opened, a random verification code pops up on the page for secondary verification; after the user enters the verification code through the dedicated keyboard, the system checks: whether the verification code is correct, whether the verification code matches the keystroke information of the user's dedicated keyboard, whether the user account and the keystroke keyboard unique identifier match, and whether the keystroke information matches the user's behavioral characteristics. After all the above checks are passed, the system opens the business platform container.

[0056] After the container is successfully opened, the user enters the specified container on the business platform container page. When entering, a random verification code pops up on the page for secondary verification. After the user enters the verification code through a dedicated keyboard, the system checks: whether the verification code is correct, whether the verification code matches the keystroke information of the user's dedicated keyboard, whether the user account and the unique identifier of the keystroke keyboard match, and whether the keystroke information matches the user's behavioral characteristics. After all the above checks are passed, the remote desktop connection is opened, and the user enters the container remotely to perform business operations.

[0057] Among them: the container may have multiple entries. For example, when the operator leaves the room midway, he is required to exit the container and log out. When the security detection module detects that there is no operation for a long time, the operator will be kicked out of the container and log out. When the operator comes back to continue working, he needs to log in again and re-enter the container.

[0058] Virtual machines implement containers For some heavy-load work scenarios, such as video processing and image rendering, you can choose a virtual machine as a container to achieve a better user experience. When opening a container, the user selects the container type as a virtual machine, and specifies the required operating system and required performance parameters. The system creates the corresponding virtual machine according to the requirements and imports the authorized data. At the same time, it creates an independent security detection module for the virtual machine. After the above steps are completed, the container is successfully opened. The container adopts a policy of easy entry and strict exit. After remote access to the container, the user can freely import the required work tools, but the export is limited to the work results and needs to be strictly reviewed. Specifically, the user selects the content to be exported through the built-in export tool (the only export channel) of the container. The exported content will be temporarily stored in the review area. After the security personnel manually review it, the user can download his own exported content in the external link. The user's operation process in the container is monitored by the supporting independent security detection module throughout the process, including real-time detection of various system logs of the virtual machine operating system, network usage, user operation video, etc. After the user's work is completed, the system destroys the virtual machine and archives information such as system logs, network usage, user operation video, and security module inspection process for audit.

[0059] Dedicated Servers Implement Containers For some work scenarios that require high-performance computing or special hardware support, a dedicated server can be selected as a container implementation. When opening a container, the user selects the container type as a dedicated server and specifies the required operating system and required performance parameters. The system selects an idle dedicated server that meets the requirements and imports the authorized data. At the same time, it creates an independent security detection module for the dedicated server. After the above steps are completed, the container is successfully opened.

[0060] The container adopts a policy of easy entry and strict exit. After remotely accessing the container, users can freely import the required work tools, but the export is limited to work results and needs to be strictly reviewed. Specifically, users select the content to be exported through the built-in export tool of the container (the only exit channel). The exported content will be temporarily stored in the review area. After manual review by security personnel, users can download their exported content in the external link.

[0061] The user's operation process in the container is fully monitored by the supporting independent security detection module, including real-time detection of various system logs of the dedicated server operating system, network usage, user operation videos, etc. After the user's work is completed, the system image restores the dedicated server to its initial state and sets the dedicated server to an idle state. At the same time, the system logs, network usage, user operation videos, security module inspection process and other information are archived for auditing.

[0062] Input device integration 2.1 Keyboard Integration For key operations of users on the business platform, the system will pop up a random verification code for secondary verification. After the user enters the verification code through the dedicated keyboard, the system checks whether the verification code is correct, whether the verification code matches the keystroke information of the user's dedicated keyboard, whether the user account and the unique identifier of the keystroke keyboard match, and whether the keystroke information matches the user's behavioral characteristics. The operation will be triggered only after all the above checks are passed. Typical key operations include logging in, opening a container, entering a container, and exporting work results.

[0063] Mouse integration When doing data annotation work (for example, before AI training for target recognition, you need to manually select the target object in the picture and then feed it to the algorithm for model training), the user mainly uses the mouse to draw lines or draw frames on the picture. In this case, the keyboard may not be pressed for a long time. At this time, the mouse keystroke information is needed for security protection. Specifically, when the user is annotating, the system continuously checks whether the user account and the keystroke mouse unique identifier match, whether the keystroke information matches the user's behavior characteristics, and when there is no keystroke information on the keyboard and mouse, it is determined that the operator has left the seat, and the system exits the container and logs out.

[0064] Keyboard and mouse data acquisition device The implementation of keyboard and mouse data acquisition devices is divided into two types: hardware data acquisition devices and software data acquisition devices. Under normal circumstances, hardware data acquisition devices should be used, but hardware data acquisition devices have requirements for wireless networks, such as 4G or 5G network coverage. If the conditions are not met, software data acquisition devices can be used instead.

[0065] Hardware data acquisition device The hardware data acquisition device mainly consists of a USB input interface, a USB output interface, a keyboard and mouse signal acquisition module, and a signal transmission module (all 3 / 4 / 5G standards are acceptable). When in use, connect the keyboard or mouse to the USB input interface, and connect the user terminal PC to the USB output interface. When the keyboard and mouse keystroke information passes through the device, it will be bypassed by the keyboard and mouse signal acquisition module and transmitted to the remote keyboard and mouse signal storage server through the wireless network via the signal transmission module in the form of MQTT protocol (widely used in the Internet of Things). The unique keyboard and mouse identifier is generated by the hardware data acquisition device and will be transmitted to the remote keyboard and mouse signal storage server along with the keystroke information.

[0066] Software Data Acquisition Device The software data collection device is an independent program installed on the user terminal PC. When the keystroke information of the keyboard and mouse enters the user terminal PC, it will be bypassed by the collection program and transmitted to the remote keyboard and mouse signal storage server through an independent proprietary network. The unique keyboard and mouse identifier is generated by the collection program based on the hardware information and will be transmitted to the remote keyboard and mouse signal storage server along with the keystroke information.

Claims

1. A system in which data can be used in a visible and safe manner, characterized in that: It includes a user dual-channel input device, an input information storage server, a business operation module, and a security detection module; The user dual-channel input device is used to output the user's operation in two channels, one channel connected to the business operation server / container, and the other channel connected to the input information storage server; The input information storage server is used to store the unique identification number of the input device, input operation information, and time information for subsequent security detection; The business operation module includes a business operation server / container and business software for using data; the business operation server / container is used to provide a closed data use environment to ensure that the data loaded in the closed data use environment is closed and safely used; the business operation server is started in advance for standby or according to user needs, and the business operation container is pulled up according to user needs; The business software is configured or customized according to the scenario of data use to process and use the data; The security detection module is used to check whether the user's behavior in the business operation module is normal during the operation of the security detection program. When an abnormality is found, the corresponding response operation is performed according to the security policy configuration.

2. A system for visible and safe use of data according to claim 1, characterized in that: The user dual-channel input device is at least one of the user's local terminal, keyboard, mouse, screen touch device, and input splitter; the input operation information stored in the input information storage server includes but is not limited to keyboard / mouse keystroke information and touch screen / touch input device click information.

3. A system for visible and safe use of data according to claim 1, characterized in that: When the business operation container is pulled up or the business operation server is started, the business software and the encrypted data to be used are loaded, and the data to be used is decrypted after obtaining the authorization key. The user remotely accesses the data closed use environment to use the data. During the entire use process, the data is prohibited from leaving the data closed use environment. Each time the business operation server / container works, a corresponding business operation log will be generated.

4. A system for visible and safe use of data according to claim 1, characterized in that: When the credibility of the parties using the data meets the requirements, after the business operation container / server is started, users can use the data by remotely accessing the closed data usage environment.

5. The system for making data visible and safe to use according to claim 1, characterized in that: A closed data environment is achieved through dedicated hardware servers in conjunction with security measures; or virtualization technology is used in conjunction with security measures, or containerization technology is used in conjunction with security measures to build a closed data environment; wherein, the dedicated hardware server refers to a server environment established according to the data usage scenario, using the necessary security network configuration combined with security protection software and hardware to prohibit data from being transmitted out of the server at will.

6. A system for visible and safe use of data according to claim 1, characterized in that: The inspection is to check whether the user's behavior in the business operation module is normal, and the inspection strategy is selected or customized according to business needs. The inspection strategy includes one or more of whether the user account and the dual-way input device identifier match, whether the user performs illegal operations, whether the user operation and input information match, and whether the user's keystroke behavior characteristics match the user himself; the inspection strategy adopts fixed-point inspection, regular inspection, fixed-point and regular inspection, or continuous inspection according to business needs.

7. A system for visible and safe use of data according to claim 1, characterized in that: The response operations configured by the security detection module include but are not limited to immediately interrupting user operations, destroying data, prompting abnormalities, requesting three-party confirmation, and sending early warning information.

8. A method for implementing a system for visible and safe use of data based on any one of claims 1 to 7, characterized in that The following steps are involved: Step 1: Build a business operation server / container to provide a closed data usage environment. Data users obtain authorization and load authorized data into the closed data usage environment. Step 2: The user uses the input information generated by the dual-channel input device to access the business operation server / container through the first channel to perform data operation; when the user performs business operation, the dual-channel input device transmits the input information to the input information storage server through the second channel for storage; Step 3: The security detection module extracts the user's usage operation behavior in the business operation module and the input information in the input information storage server, and checks the extracted usage operation behavior and input information according to the set inspection strategy. If an abnormality is found, a response operation is taken according to the configuration.

9. The method for making data visible and safe to use according to claim 8, characterized in that: In step 3, the use operation behavior in the business operation module is extracted by monitoring various system logs or by parsing the received business operation data packets.

10. The method for making data visible and safe to use according to claim 8, characterized in that: The security detection module at least includes archiving the user's operation behavior, input information and information recorded during the inspection process to provide data support for auditing and backtracking; Input the information stored in the information storage server, combine it with the business operation log, review the illegal operations in the data usage process, and trace the user's violation responsibility.

Images