Real-time joint decision-making system and method based on privacy protection
By configuring the data source in the joint decision-making system and parsing it into a directed acyclic graph, combining secret sharing mode and secure operator technology, the problems of data transmission security and real-time decision-making are solved, and data privacy protection and efficient decision-making flow are achieved quickly Orchestrate and execute.
Patent Information
- Application Number
- CN202411069976.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-06
- Publication Date
- 2024-11-08
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The existing joint decision-making system cannot protect data security during data transmission, and the decision-making flow of federated learning cannot achieve real-time decision-making, resulting in data privacy being unable to guarantee and decision-making efficiency.
Using a real-time joint decision-making system based on privacy protection, by obtaining the module to configure the data source, the storage module converts the data source into nodes and stores it, the parsing module parses the nodes into directed acyclic graphs, and uses secret sharing during the decision-making process. Mode and security operator technology protect external data privacy and realize real-time decision-making.
It realizes the separation of data source services and decision templates. Service changes do not require modification of back-end codes. It can quickly orchestrate decision flows, protect external data privacy, make real-time decisions, save resource consumption, and quickly return decision results.
Smart Images

Figure CN118916914A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of computer technology, and in particular to a real-time joint decision-making system and method based on privacy protection. Background Art
[0002] At present, many small enterprises build their own approval processes, that is, they add designated process personnel to the nodes according to the internal process, and promote the flow of approval by having process personnel manually complete tasks and then approve them. However, this is prone to problems such as untimely approval and waste of manpower.
[0003] Going one step further, there is the emergence of automatic decision flow (i.e. rule flow), which automatically judges and reviews the input data based on the approval flow through the configured evaluation criteria, greatly saving manpower, but the data is often limited to one party. On this basis, there is joint decision-making, which can create more complex decision flows by introducing external data sources and performing associated query and judgment on data from multiple data sources through data primary keys (such as user ID numbers).
[0004] The more mature joint decision-making rule engines on the market are:
[0005] jvs: You can customize data sources, variables, canvases, and decision types, but data is transmitted in plain text and cannot protect data security.
[0006] Other decision flows based on federated learning: data privacy can be protected, but federated learning makes decisions offline and cannot make real-time decisions.
[0007] In view of the above technical problems, the present invention proposes a real-time joint decision-making system and method based on privacy protection. Summary of the invention
[0008] The purpose of the present invention is to provide a real-time joint decision-making system and method based on privacy protection in view of the defects of the prior art.
[0009] In order to achieve the above purpose, the present invention adopts the following technical solutions:
[0010] A real-time joint decision-making system based on privacy protection, comprising:
[0011] The acquisition module is used to obtain the required data source and configure the acquired data source;
[0012] The storage module is used to convert the configured data source into the corresponding node and store it in the database;
[0013] The parsing module is used to parse the stored nodes to obtain the decision flow of the joint decision.
[0014] Furthermore, the data sources in the acquisition module include local assets, API data sources, and external assets.
[0015] Furthermore, the external assets are privacy-protected external assets.
[0016] Furthermore, after configuring the acquired data source in the acquisition module, the step further includes: packaging the configured data source into a corresponding variable.
[0017] Furthermore, the parsing module performs parsing processing on the stored nodes specifically as follows: parsing the stored nodes into a directed acyclic graph.
[0018] Furthermore, after parsing the stored nodes into a directed acyclic graph, the method further includes: packaging the directed acyclic graph.
[0019] Accordingly, a real-time joint decision-making method based on privacy protection is also provided, including:
[0020] S1. Obtain the required data source and configure the obtained data source;
[0021] S2. Convert the configured data source into the corresponding node and store it in the database;
[0022] S3. Analyze and process the stored nodes to obtain the decision flow of the joint decision.
[0023] Furthermore, the step S1 also includes: packaging the configured data source into a corresponding variable.
[0024] Furthermore, the parsing process of the stored nodes in step S3 is specifically: parsing the stored nodes into a directed acyclic graph.
[0025] Furthermore, the step S3 also includes packaging the directed acyclic graph.
[0026] Compared with the prior art, the present invention has the following beneficial effects:
[0027] 1. You can quickly arrange the decision flow by simply dragging and dropping the canvas;
[0028] 2. The data source service is separated from the decision template. Service changes do not require modifying the backend code, only adjusting the variable configuration;
[0029] 3. Protect external data privacy throughout the decision-making process, and the data is available but not visible;
[0030] 4. Real-time decision-making, saving resource consumption during calculation and quickly returning decision results. BRIEF DESCRIPTION OF THE DRAWINGS
[0031] Figure 1 This is a structural diagram of a real-time joint decision-making system based on privacy protection provided in Example 1;
[0032] Figure 2 This is a schematic diagram of the API data source interface provided in Example 1;
[0033] Figure 3 This is a schematic diagram of the variable management area interface provided in Example 1;
[0034] Figure 4 This is a schematic diagram of rule configuration provided in Example 1. DETAILED DESCRIPTION
[0035] The following describes the embodiments of the present invention by specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the following embodiments and features in the embodiments can be combined with each other without conflict.
[0036] The purpose of the present invention is to provide a real-time joint decision-making system and method based on privacy protection in view of the defects of the prior art.
[0037] Embodiment 1
[0038] This embodiment provides a real-time joint decision-making system based on privacy protection, such as Figure 1 As shown, including:
[0039] The acquisition module is used to obtain the required data source and configure the acquired data source;
[0040] The storage module is used to convert the configured data source into the corresponding node and store it in the database;
[0041] The parsing module is used to parse the stored nodes to obtain the decision flow of the joint decision.
[0042] In this embodiment, the real-time joint decision-making system based on privacy protection is divided into a front-end, a business control layer, a stream engine layer, and a data service layer according to the framework.
[0043] The data service layer is used to provide data sources for federated computing, including API data sources that can be used in plain text, local assets, and external assets that require privacy protection.
[0044] The front-end page mainly includes the data source management area, variable management area and rule configuration area in the order of configuration. Users configure the data source in the data source management area, then package the configured data source into variables in the variable management area, and finally use the configured variables in the rule management area to form rule nodes.
[0045] The data source management area includes an acquisition module for introducing data sources, acquiring API data sources, local assets, external assets and other data sources provided by the data service layer, and configuring the acquired data sources.
[0046] In this embodiment, the configuration of the acquired API data source, local assets, and external assets is specifically as follows:
[0047] The API data source encapsulates the HTTP protocol and uses the POST / GET format to query data. The API interface diagram is as follows: Figure 2 shown.
[0048] Our assets are locally stored data and can be saved in various forms of databases, or directly uploaded to a designated file library using FTP, local files, etc.
[0049] External assets are authorized by external services. You can get the metadata of the assets, that is, which fields the assets are composed of and what types they are, but you cannot get the specific data information. In actual use, the security operator is used to determine the specified attribute value of the specified primary key data.
[0050] The variable management area is used to package the configured data source into variables, such as Figure 3 As shown, the packaging method includes the selection of variable type, the configuration of non-assigned variables, the definition of variable names, the description of variables, etc.
[0051] The rule configuration area includes the variable configuration area and the condition configuration area; you can pre-configure node components (start, end, rule, branch, assignment, etc.) by dragging and dropping, and assemble the required process on the canvas as needed. The rule nodes involved in the process are used to configure single-condition judgments, the branch nodes involved in the process are used to judge multiple conditions, and the assignments involved in the process are used to assign values to the specified variables before the end, so as to facilitate the distinction of specific results. After configuring the process nodes, you can click to select the specified node, first introduce the previously configured variables in the variable configuration area of the specified type node, and then use the introduced variables in the condition configuration area to construct the judgment conditions, such as a is greater than or equal to b, where a is the value of the variable output and b is the configured value. "Greater than or equal to" belongs to the judgment condition, and the specific judgment types include: contain, not contain, equal, not equal, greater than, greater than or equal to, less than, less than or equal to, etc. Configure the subsequent actions of the judgment result. For example, if the judgment is true, go to the next node through this node, or end the process directly. You can also configure the value of the specified assignment variable as a parameter in the assignment node. After configuration is completed, save the configuration condition table into the database. The schematic diagram of the rule configuration area is as follows: Figure 4 shown.
[0052] The business control layer interacts with the front end. The business control layer includes a storage module, which is used to convert the configured data source into corresponding nodes and store them in the database, and then provide an execution interface to call the underlying decision flow engine.
[0053] The database of this embodiment stores the nodes, connection data, conditions, results and other contents in the configuration process.
[0054] The stream engine layer includes a parsing module, which is used to parse the stored nodes to obtain a decision flow for joint decision making.
[0055] The processing module of this embodiment is responsible for the parsing, verification and processing of the entire stream, which specifically includes:
[0056] DSL parsing component, used to parse the data structure of nodes and connections stored in the database into a DAG (directed acyclic graph) structure;
[0057] DAG verification component is used to verify whether the entire process is complete, including whether the DAG structure is correct and complete, and whether the nodes and conditions are configured reasonably.
[0058] DAG packaging component is used to package the executable DAG structure so that it has a data structure that can be used by the execution engine, including node assembly, condition assembly, variable assembly, parameter assembly, etc.
[0059] The parameter conversion component is used to convert the incoming parameters into the actual parameters required for conditional judgment in the process through the acquisition module. The API data source and local assets can directly obtain data, while the external data source is in the subsequent conditional judgment component. Through the abstract external federated interface, the secret sharing mode is adopted to obtain the judgment result by transmitting the variable name, comparison symbol and judgment threshold, thereby playing a role in privacy protection;
[0060] Process scheduler, used to advance the entire process;
[0061] The component container is a collection of various components, used for task processing of various types of components, and distributed according to type;
[0062] Conditional judgment component, used to judge specific conditions in the process.
[0063] The front-end, business control layer, flow engine layer and data service layer of the utility model are separated. Business modifications will not affect the logic of the engine, and modifications to the data source will not affect the determination of the upper-layer rules. Decoupling is achieved in a layered manner. For reusable data sources, the concept of variables is abstracted to achieve rapid reuse, without the need to reconfigure the required data sources and parameters each time a new decision-making process is built.
[0064] To support real-time decision-making, the parameters are grouped in the parameter conversion component, and the conversion parameters are queried in batches in the form of multiple tasks. Compared with the lazy loading mode that queries only when needed, this method can improve the operation efficiency when facing complex processes that require a large number of parameters.
[0065] This embodiment also performs data comparison at the external data source end through dedicated services, secret sharing and security operator technology, which not only avoids data leakage caused by the circulation of plaintext data, but also realizes the normal use of data.
[0066] Embodiment 2
[0067] This embodiment provides a real-time joint decision-making method based on privacy protection, including:
[0068] S1. Obtain the required data source and configure the obtained data source;
[0069] S2. Convert the configured data source into the corresponding node and store it in the database;
[0070] S3. Analyze and process the stored nodes to obtain the decision flow of the joint decision.
[0071] Furthermore, the step S1 also includes: packaging the configured data source into a corresponding variable.
[0072] Furthermore, the parsing process of the stored nodes in step S3 is specifically: parsing the stored nodes into a directed acyclic graph.
[0073] Furthermore, the step S3 also includes packaging the directed acyclic graph.
[0074] It should be noted that the real-time joint decision-making method based on privacy protection provided in this embodiment corresponds to Embodiment 1 and will not be elaborated here.
[0075] This embodiment has the following beneficial effects:
[0076] 1. You can quickly arrange the decision flow by simply dragging and dropping the canvas;
[0077] 2. The data source service is separated from the decision template. Service changes do not require modifying the backend code, only adjusting the variable configuration;
[0078] 3. Protect external data privacy throughout the decision-making process, and the data is available but not visible;
[0079] 4. Real-time decision-making, saving resource consumption during calculation and quickly returning decision results.
[0080] Note that the above are only preferred embodiments of the present invention and the technical principles used. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and that various obvious changes, readjustments and substitutions can be made by those skilled in the art without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in more detail through the above embodiments, the present invention is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present invention, and the scope of the present invention is determined by the scope of the appended claims.
Claims
1. A real-time joint decision-making system based on privacy protection, characterized in that: include: The acquisition module is used to obtain the required data source and configure the acquired data source; The storage module is used to convert the configured data source into the corresponding node and store it in the database; The parsing module is used to parse the stored nodes to obtain the decision flow of the joint decision.
2. The real-time joint decision-making system based on privacy protection according to claim 1, characterized in that: The data sources in the acquisition module include local assets, API data sources, and external assets.
3. The real-time joint decision-making system based on privacy protection according to claim 2, characterized in that: The external assets are privacy-protected external assets.
4. The real-time joint decision-making system based on privacy protection according to claim 3 is characterized in that: After configuring the acquired data source in the acquisition module, the method further includes: packaging the configured data source into a corresponding variable.
5. The real-time joint decision-making system based on privacy protection according to claim 4 is characterized in that: The parsing module performs parsing processing on the stored nodes specifically as follows: parsing the stored nodes into a directed acyclic graph.
6. The real-time joint decision-making system based on privacy protection according to claim 5, characterized in that: After parsing the stored nodes into a directed acyclic graph, the method further includes: packaging the directed acyclic graph.
7. A real-time joint decision-making method based on privacy protection, characterized in that: include: S1. Obtain the required data source and configure the obtained data source; S2. Convert the configured data source into the corresponding node and store it in the database; S3. Analyze and process the stored nodes to obtain the decision flow of the joint decision.
8. The real-time joint decision-making method based on privacy protection according to claim 7 is characterized in that: The step S1 also includes: packaging the configured data source into a corresponding variable.
9. The real-time joint decision-making method based on privacy protection according to claim 8, characterized in that: The step S3 of parsing the stored nodes specifically includes: parsing the stored nodes into a directed acyclic graph.
10. The real-time joint decision-making method based on privacy protection according to claim 9, characterized in that: The step S3 also includes packaging the directed acyclic graph.
Citation Information
Patent Citations
Federal learning system based on block chain
CN113420335A
Risk control system of financing lease risk control rule
CN116957759A
Data quality evaluation method and system for original data non-out-of-domain scene
CN117807626A
United rule product design method and system based on privacy protection
CN118134616A
