Method and system for realizing data sharing and opening based on large language model
Through a large language model, the data source tool plug-in package desensitization and data calculation process control is solved, and the problem of insufficient security control in public sector data sharing is solved, and data security and convenience of use is achieved. Improvement.
Patent Information
- Application Number
- CN202411091539.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-09
- Publication Date
- 2024-12-06
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
In the prior art, public departments lack security controls when sharing data, resulting in sensitive data leakage and high threshold for data use. They are only applicable to structured data and cannot effectively monitor the acquisition, use and dissemination of data demanders.
Using a large language model to encapsulate and desensitize the data source tool plug-in, control the data calculation process, and ensure the security of data demanders' data acquisition, while reducing the threshold for data use and enriching data processing types. .
It realizes security control of the data demander, prevents sensitive data leakage, lowers the threshold for data use, and supports multiple data types to ensure that data is calculated within the provider server and avoids data leakage.
Smart Images

Figure CN119089488A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of public service technology, and specifically to a method and system for realizing data sharing and openness based on a large language model. Background Art
[0002] At present, when the government and other public departments provide data services to the outside world as data providers, generally after the data demander obtains authorization, the data provider's business application system provides the corresponding interface or library table and pushes it to the data demander through the data sharing and exchange platform. The data demander can obtain these sensitive user data based on the interface or library table.
[0003] This method lacks the necessary security control capabilities and does not require authorization from specific users involved in the data. The data demander can obtain the user's sensitive information through interfaces or database tables at will only after obtaining authorization from the data provider. At this time, the use and dissemination of sensitive data by the data demander cannot be controlled, which may cause the leakage of sensitive data and security issues. This method has a relatively high threshold for data use by the data demander. The data demander must deploy a business application system to use data through interfaces or database tables. This method processes limited data types and is usually only applicable to structured data.
[0004] Therefore, a method and system for achieving data sharing and openness based on a large language model has become an urgent problem to be solved. Summary of the invention
[0005] In response to the security problem in the prior art that data providers are unable to monitor the acquisition, use and dissemination of data by data demanders, the present invention provides a method and system for realizing data sharing and openness based on a large language model. The large language computing model is used to encapsulate and desensitize data source tool plug-ins, control the data computing process, verify and review data input and output, and monitor the full log of the interaction process to achieve security control over data acquisition by data demanders, thereby achieving the purpose of ensuring data security, while lowering the threshold for data use and enriching data processing types.
[0006] In order to solve the above technical problems, the present invention provides a technical solution: a method for realizing data sharing and openness based on a large language model, which involves the following three aspects:
[0007] (A) Data source tool plug-in encapsulation and desensitization: Analyze, aggregate, extract, and desensitize the data source of the data provider, encapsulate the same type of data into the same tool plug-in, and process the data in the tool plug-in according to the requirements of the data demander to generate a data source tool plug-in for use by the data demander;
[0008] (B) Data computing process deployment application: Design a computing process based on the results that the data demander wants to obtain based on the data, and deploy it to the data provider's server; when the data demander initiates an access request to the data provider, the data computing process is retrieved, and the retrieved computing process is used to process the data, and then the computing results are returned to the data demander;
[0009] (C) Data input and output security control: including input control and output control. The data provider determines whether the data requester has access rights through data input control, and the data provider verifies and reviews whether the output results of the calculation process meet the requirements through data output control.
[0010] Optionally, according to the results that the data demander wants to obtain based on the data, the designed calculation process includes data input information, large language model processing information and calculation result data output information;
[0011] According to the results that the data demander wants to obtain based on the data, the designed calculation process is deployed to the data provider's server. The data provider sets the large language model based on the data input information of the calculation process and automatically specifies the data source tool plug-in that the calculation process needs to call according to the context environment.
[0012] Optionally, in the data source tool plug-in packaging stage, multiple data source tool plug-ins are obtained, and the multiple data source tool plug-ins are added to the tool plug-in list;
[0013] When the data demander initiates an access request to the data provider, the data provider calls the calculation process related to the data demander. The calculation process further selects the data source tool plug-in related to it in the list for calculation and returns the calculation result to the data demander.
[0014] Further optionally, the data provider determines whether the data requester has access rights through data input control, specifically including:
[0015] Determine whether the interface key of the data demander is correct through data input control;
[0016] Determine whether the user ID of the data requester has been registered with the data provider through data input control;
[0017] Through data input control, it is determined whether the data classification required by the data demander is within a reasonable range.
[0018] Optionally, the data provider may verify whether the output results of the calculation process meet the requirements through data export control, including:
[0019] Verify and review the output results of the calculation process through data export control to see whether they meet the standard format information required for the calculation results;
[0020] Data export control is used to verify whether the output results of the calculation process meet the sensitive content review requirements.
[0021] Secondly, the present invention provides a system for realizing data sharing and opening based on a large language model, and the system comprises:
[0022] The data source tool plug-in encapsulation module is used to analyze, summarize, extract, and desensitize the data source of the data provider, encapsulate the same type of data into the same tool plug-in, and process the data in the tool plug-in according to the requirements of the data demander to generate a data source tool plug-in for use by the data demander;
[0023] The data calculation process module is used to assist data demanders in designing calculation processes based on the results they want to obtain based on the data, and deploy them to the servers of data providers;
[0024] The large language model application module is used to assist the data provider in calling the calculation process according to the data requirements of the access request. The calculation process selects the data source tool plug-in for calculation and returns the calculation results to the data demander.
[0025] The data input security control module is used to control data input. The data provider uses input control to determine whether the data demander's interface key is correct and whether it has access rights, and whether the required data classification is within a reasonable range.
[0026] The data output security control module is used to control data output. The data provider verifies through output control whether the output results of the calculation process conform to the required standard format information and whether they meet the sensitive content review requirements.
[0027] Optionally, when the data provider designs the calculation process through the data calculation process module, the calculation process contains data input and output information and calculation result information;
[0028] After the data provider deploys the calculation process to the data provider's server through the data calculation process module, the data provider specifies the data source tool plug-in that the calculation process needs to call through the large language model application module based on the data entry information of the calculation process.
[0029] Optionally, the data source tool plug-in encapsulation module involved generates multiple data source tool plug-ins, and then adds the multiple data source tool plug-ins to the tool plug-in list;
[0030] When the data demander initiates an access request to the data provider, the data provider calls the calculation process related to the data demander. The calculation process further selects the data source tool plug-in related to it in the list for calculation and returns the calculation result to the data demander.
[0031] Optionally, the data provider determines whether the data requester has access rights through data input security control, including:
[0032] Determine whether the interface key of the data demander is correct through the input control of the data input security control module;
[0033] Determine whether the user ID of the data requester has been registered with the data provider through the input control of the data input security control module;
[0034] The input control of the data input security control module determines whether the data classification required by the data demander is within a reasonable range.
[0035] Optionally, the data provider may verify whether the output results of the calculation process meet the requirements through data export control, including:
[0036] Check and verify whether the output results of the calculation process meet the standard format information required by the calculation results through the data output control of the data output security control module;
[0037] The data output control of the data output security control module verifies whether the output results of the calculation process meet the sensitive content review requirements.
[0038] The advantages of the present invention compared with the prior art are:
[0039] 1. The present invention implements security control over data acquisition by data demanders by encapsulating and desensitizing data source tool plug-ins, controlling data calculation processes, verifying and reviewing data input and output, and monitoring the entire log of the interaction process, thereby achieving the purpose of ensuring data security, while lowering the threshold for data use and enriching data processing types.
[0040] 2. The present invention designs a calculation process based on the needs of the data demander, and embeds the designed calculation process into the server of the data provider. All calculations are performed in the server of the data provider, thereby preventing data from being leaked, facilitating management and ensuring information security. BRIEF DESCRIPTION OF THE DRAWINGS
[0041] Figure 1 It is a flow chart of a method for realizing data sharing and opening based on a large language model of the present invention.
[0042] Figure 2 It is a system block diagram of a system for realizing data sharing and opening based on a large language model according to the present invention. DETAILED DESCRIPTION
[0043] Various exemplary embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be noted that unless otherwise specifically stated, the relative arrangement of components and steps, numerical expressions and numerical values set forth in these embodiments do not limit the scope of the present invention.
[0044] The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the invention, its application, or uses.
[0045] Technologies, methods, and equipment known to ordinary technicians in the relevant art may not be discussed in detail, but where appropriate, the technologies, methods, and equipment should be considered as part of the specification.
[0046] In all examples shown and discussed herein, any specific values should be interpreted as merely exemplary and not as limiting. Therefore, other examples of the exemplary embodiments may have different values.
[0047] Combined with Figure 1-2 The specific implementation process of the method and system for realizing data sharing and opening based on a large language model of the present invention is as follows:
[0048] Embodiment 1:
[0049] Combined with Figure 1 This embodiment proposes a method for realizing data sharing and opening based on a large language model. The method involves the following three aspects:
[0050] (A) Data source tool plug-in encapsulation and desensitization: Analyze, aggregate, extract, and desensitize the data source of the data provider, encapsulate the same type of data into the same tool plug-in, and process the data in the tool plug-in according to the requirements of the data demander to generate a data source tool plug-in for use by the data demander;
[0051] In this stage, multiple data source tool plug-ins will be obtained and added to the plug-in list.
[0052] (B) Data computing process deployment application: Design a computing process based on the results that the data demander wants to obtain based on the data, and deploy it to the data provider's server; when the data demander initiates an access request to the data provider, the data computing process is retrieved, and the retrieved computing process is used to process the data, and then the computing results are returned to the data demander;
[0053] In this stage, according to the results that the data demander wants to obtain based on the data, the designed calculation process includes data input information, large language model processing information and calculation result data output information.
[0054] In this stage, according to the results that the data demander wants to obtain based on the data, the designed calculation process is deployed to the data provider's server. The data provider sets the large language model based on the data input information of the calculation process to automatically specify the data source tool plug-in that the calculation process needs to call according to the context environment.
[0055] When the data demander initiates an access request to the data provider, the data provider calls the calculation process related to the data demander. The calculation process further selects the data source tool plug-in related to it in the list for calculation and returns the calculation result to the data demander.
[0056] (C) Data input and output security control: including input control and output control. The data provider determines whether the data requester has access rights through data input control, and the data provider verifies and reviews whether the output results of the calculation process meet the requirements through data output control.
[0057] In this embodiment, the data provider determines whether the data requester has access rights through data input control, specifically including:
[0058] Determine whether the interface key of the data demander is correct through data input control;
[0059] Determine whether the user ID of the data requester has been registered with the data provider through data input control;
[0060] Data input control is used to determine whether the data classification required by the data demander is within a reasonable range.
[0061] In this embodiment, the data provider verifies whether the output results of the calculation process meet the requirements through data export control, specifically including:
[0062] Verify and review the output results of the calculation process through data export control to see whether they meet the standard format information required for the calculation results;
[0063] Data export control is used to verify whether the output results of the calculation process meet the sensitive content review requirements.
[0064] Based on the method of this embodiment, when the data demander initiates an access request to the data provider, the data provider first determines whether the data demander has access rights through data input control. If the data demander has access rights, the data provider calls the calculation process related to the data demander. The calculation process relies on the large language model to further select the data source tool plug-in for calculation. After the calculation process outputs the calculation result, the data provider verifies and examines whether the output result of the calculation process meets the standard format and sensitive content requirements through data output control. If it meets the requirements, the output result of the calculation process will be returned to the data demander.
[0065] Embodiment 2:
[0066] Combined with Figure 2 This embodiment proposes a system for realizing data sharing and opening based on a large language model, and the system includes:
[0067] The data source tool plug-in encapsulation module is used to analyze, summarize, extract, and desensitize the data source of the data provider, encapsulate the same type of data into the same tool plug-in, and process the data in the tool plug-in according to the requirements of the data demander to generate a data source tool plug-in used by the calculation process related to the data demander;
[0068] The data calculation process module is used to assist data demanders in designing calculation processes based on the results they want to obtain based on the data, and deploy them to the servers of data providers;
[0069] The large language model application module is used to assist the data provider in calling the data calculation process according to the data requirements of the access request. The calculation process selects the data source tool plug-in module for calculation and returns the calculation results to the data demander.
[0070] The data input security control module is used to control data input. The data provider uses input control to determine whether the data demander's interface key is correct and whether it has access rights, and whether the required data classification is within a reasonable range.
[0071] The data output security control module is used to control data output. The data provider verifies through output control whether the output results of the calculation process conform to the required standard format information and whether they meet the sensitive content review requirements.
[0072] In this embodiment, when designing a calculation process through a data calculation process module according to the needs of the data demander, the calculation process module contains data input information and data calculation result output information. After the data provider deploys the calculation process module to the server, the data provider specifies the data source tool plug-in that the calculation process needs to call through the large language model application module according to the data input information of the calculation process module.
[0073] In this embodiment, the data source tool plug-in encapsulation module extracts multiple data source tool plug-ins, and then adds the multiple data source tool plug-ins to the plug-in list. When the data demander initiates an access request to the data provider, the data provider calls the calculation process designed by the data demander, and the calculation process selects the data source tool plug-in related to it in the list for calculation, and returns the calculation result to the data demander.
[0074] In this embodiment, based on the data input security control module, the data provider determines whether the data demander has access rights through data input control, specifically including:
[0075] The data input security control module determines whether the interface key of the data demander is correct;
[0076] The data input security control module determines whether the user ID of the data requester has been registered with the data provider;
[0077] The data input security control module is used to determine whether the data classification required by the data demander is within a reasonable range.
[0078] In this embodiment, based on the data output security control module, the data provider verifies whether the output result of the calculation process meets the requirements through data output control, which specifically includes:
[0079] Verify and review the output results of the calculation process through the data output security control module to see if they meet the standard format information required for the calculation results;
[0080] The data output security control module is used to verify whether the output results of the calculation process meet the sensitive content review requirements.
[0081] In summary, the method and system for realizing data sharing and openness based on a large language model of the present invention realizes security control over data acquisition by data demanders through encapsulation and desensitization of data source tool plug-ins, control of data calculation processes, verification and review of data input and output, and full log monitoring of the interaction process, thereby achieving the purpose of ensuring data security, while lowering the threshold for data use and enriching data processing types.
[0082] The present invention and its embodiments are described above, and such description is not restrictive. The drawings show only one embodiment of the present invention, and the actual structure is not limited thereto. In short, if ordinary technicians in the field are inspired by it, without departing from the purpose of the invention, they can design a structure and embodiment similar to the technical solution without creativity, which should belong to the protection scope of the present invention.
Claims
1. A method for realizing data sharing and opening based on a large language model, characterized in that: include S1. Data source tool plug-in encapsulation and desensitization Analyze, summarize, extract, and desensitize the data source of the data provider, encapsulate the same type of data into the same tool plug-in, and process the data in the tool plug-in according to the requirements of the data demander to generate a data source tool plug-in for the data demander to use; S2. Data computing process deployment application Design the calculation process based on the results that the data demander wants to obtain based on the data, and deploy it to the server of the data provider; When the data demander initiates an access request to the data provider, the data calculation process is retrieved, and the data is processed using the retrieved calculation process, and then the calculation result is returned to the data demander; S3. Data input and output security control The data provider determines whether the data demander has access rights through data input control, and the data provider verifies whether the output results of the calculation process meet the requirements through data output control.
2. The method for realizing data sharing and opening based on a large language model according to claim 1, characterized in that: In S2, according to the results that the data demander wants to obtain based on the data, the designed calculation process includes data input information, large language model processing information and calculation result data output information; According to the results that the data demander wants to obtain based on the data, the designed calculation process is deployed to the data provider's server. The data provider sets the large language model based on the data input information of the calculation process and automatically specifies the data source tool plug-in that the calculation process needs to call according to the context environment.
3. The method for realizing data sharing and opening based on a large language model according to claim 2, characterized in that: In the data source tool plug-in packaging stage of S1, multiple data source tool plug-ins are obtained, and the multiple data source tool plug-ins are added to the tool plug-in list; When the data demander initiates an access request to the data provider, the data provider calls the calculation process related to the data demander. The calculation process further selects the data source tool plug-in related to it in the list for calculation and returns the calculation result to the data demander.
4. The method for realizing data sharing and opening based on a large language model according to claim 3, characterized in that: In S3, the data provider determines whether the data demander has access rights through data input control, including: Determine whether the interface key of the data demander is correct through data input control; Determine whether the user ID of the data requester has been registered with the data provider through data input control; Through data input control, it is determined whether the data classification required by the data demander is within a reasonable range.
5. The method for realizing data sharing and opening based on a large language model according to claim 4, characterized in that: In S3, the data provider verifies whether the output results of the calculation process meet the requirements through data export control, including: Verify and review the output results of the calculation process through data export control to see whether they meet the standard format information required for the calculation results; Data export control is used to verify whether the output results of the calculation process meet the sensitive content review requirements.
6. A system for realizing data sharing and opening based on a large language model, characterized in that: include The data source tool plug-in encapsulation module is used to analyze, summarize, extract, and desensitize the data source of the data provider, encapsulate the same type of data into the same tool plug-in, and process the data in the tool plug-in according to the requirements of the data demander to generate a data source tool plug-in for use by the data demander; The data calculation process module is used to assist data demanders in designing calculation processes based on the results they want to obtain based on the data, and deploy them to the servers of data providers; The large language model application module is used to assist the data provider in calling the calculation process according to the data requirements of the access request. The calculation process selects the data source tool plug-in for calculation and returns the calculation results to the data demander. The data input security control module is used to control data input. The data provider uses input control to determine whether the data demander's interface key is correct and whether they have access rights, and whether the required data classification is within a reasonable range; The data output security control module is used to control data output. The data provider verifies through output control whether the output results of the calculation process conform to the required standard format information and whether they meet the sensitive content review requirements.
7. The system for realizing data sharing and opening based on a large language model according to claim 6, characterized in that: When the data provider designs the calculation process through the data calculation process module, the calculation process contains data input and output information and calculation result information; After the data provider deploys the calculation process to the data provider's server through the data calculation process module, the data provider specifies the data source tool plug-in that the calculation process needs to call through the large language model application module based on the data entry information of the calculation process.
8. The system for realizing data sharing and opening based on a large language model according to claim 7, characterized in that: The data source tool plug-in encapsulation module generates a plurality of data source tool plug-ins, and then adds the plurality of data source tool plug-ins into a tool plug-in list; When the data demander initiates an access request to the data provider, the data provider calls the calculation process related to the data demander. The calculation process further selects the data source tool plug-in related to it in the list for calculation and returns the calculation result to the data demander.
9. The system for realizing data sharing and opening based on a large language model according to claim 8, characterized in that: The data provider determines whether the data requester has access rights through data input security control, including: Determine whether the interface key of the data demander is correct through the input control of the data input security control module; Determine whether the user ID of the data requester has been registered with the data provider through the input control of the data input security control module; The input control of the data input security control module determines whether the data classification required by the data demander is within a reasonable range.
10. The system for realizing data sharing and opening based on a large language model according to claim 9, characterized in that: The data provider verifies whether the output results of the calculation process meet the requirements through data export control, including: Check and verify whether the output results of the calculation process meet the standard format information required by the calculation results through the data output control of the data output security control module; The data output control of the data output security control module verifies whether the output results of the calculation process meet the sensitive content review requirements.
