Model creation method, apparatus, device, storage medium, and program product

Abstract

The application provides a model creation method and device, equipment, storage medium and program product, and relates to the field of financial technology. The method comprises the following steps: sending the encrypted data packet to the computing node through the encryption transmission unit; decrypting the encrypted data packet in the computing node to obtain decrypted data; opening an isolated memory to extract features of the decrypted data to obtain a fusion feature data set; and constructing a first business model in the distributed data sharing system based on the fusion feature data set. The data provider transmits local data to the computing node after encrypting the local data, and then decrypts the local data. After that, data cleaning and feature extraction are performed in the isolated memory to generate a fusion feature data set. Finally, a global optimal decision based on multi-source data is realized by solving the operational research model. The contradiction between cross-institutional credit investigation data fusion and privacy protection is solved.

Description

Technical Field

[0001] This application relates to the field of financial technology, and in particular to a model creation method, apparatus, device, storage medium, and program product. Background Technology

[0002] Credit data is fundamental to areas such as financial lending and risk management. Traditionally, credit data mainly comes from the credit records of financial institutions, but with the development of the digital economy, credit data sources have become increasingly diversified, including tax, social security, utility bills, telecommunications payments, supply chain fulfillment, and online behavior.

[0003] This data is scattered across different institutions (such as governments, banks, telecom operators, and internet platforms), forming "data silos." Meanwhile, data privacy regulations (such as the Personal Information Protection Law) impose strict requirements on data sharing, leading to significant challenges in the integration of cross-institutional credit data. On the other hand, operations research, as an applied mathematics discipline, is widely used in financial decision-making, such as credit approval, credit limit allocation, risk pricing, and portfolio optimization. Operations research methods (such as linear programming, integer programming, stochastic programming, dynamic programming, and game theory) can help financial institutions make optimal decisions under resource constraints and risk control.

[0004] However, the accuracy of these decision-making models is highly dependent on the comprehensiveness and timeliness of the input data. Due to the lack of cross-institutional data sharing mechanisms, existing operations research decisions often rely solely on data from a single institution or limited credit reports, making it difficult to achieve global optimum. Summary of the Invention

[0005] This application provides a model creation method, apparatus, device, storage medium, and program product to solve the technical problem that it is difficult to achieve global optimization with data from a single organization or limited credit reports.

[0006] In a first aspect, this application provides a model creation method, comprising: sending the encrypted data packet to the computing node through the encrypted transmission unit; decrypting the encrypted data packet in the computing node to obtain decrypted data; allocating isolated memory to extract features from the decrypted data to obtain a fused feature dataset; and constructing a first business model in the distributed data sharing system based on the fused feature dataset, wherein the first business model is implemented based on an operations research model.

[0007] In one possible implementation, sending the encrypted data packet to the computing node via the encrypted transmission unit includes: performing difference detection on the historical data of the target subject and the encrypted data packet via the data difference unit to obtain difference data; and sending the difference data to the computing node.

[0008] In one possible implementation, the step of decrypting the encrypted data packet in the computing node to obtain decrypted data includes: authenticating the data requester through the distributed identity authentication unit, tagging the authenticated data requester to obtain an authentication result; and decrypting the corresponding encrypted data packet based on a preset decryption strategy according to the authentication result to obtain the decrypted data.

[0009] In one possible implementation, before sending the encrypted data packet to the computing node via the encrypted transmission unit, the method further includes: dynamically authorizing and verifying the access policy of the data provider node through the smart contract unit to obtain the compliant data provider node; the dynamic authorization verification includes comparing the access range of the request field of the data requester with the preset field of the data provider; at the compliant data provider node, the provided data is encrypted based on a preset encryption method to obtain the encrypted data packet; the preset encryption method includes attribute-based encryption and homomorphic encryption; the attribute-based encryption is used to dynamically control data access permissions according to the attributes of the decryptor; and the homomorphic encryption is used to directly perform computation operations without decryption.

[0010] In one possible implementation, constructing a first business model in the distributed data sharing system based on the fused feature dataset includes: constructing an operations research model according to the objective function, the constraints, and the decision variables; and solving the operations research model by calling the fused feature dataset based on the distributed data sharing system to obtain the first business model.

[0011] In one possible implementation, the operations research model includes at least one of the following: a linear programming model for maximizing expected returns under capital constraints; and a stochastic programming model for handling uncertainty constraints.

[0012] In one possible implementation, the method further includes: uploading the generation process of the fused feature dataset and the first business model to the blockchain supervision node through the evidence storage unit to form an immutable audit log.

[0013] Secondly, this application provides a model creation apparatus, comprising:

[0014] The sending module is used to send the encrypted data packet to the computing node through the encrypted transmission unit;

[0015] A decryption module is used to decrypt the encrypted data packet in the computing node to obtain decrypted data;

[0016] The extraction module is used to allocate isolated memory to extract features from the decrypted data to obtain a fused feature dataset;

[0017] A construction module is used to construct a first business model in the distributed data sharing system based on the fused feature dataset. The first business model is implemented based on an operations research model.

[0018] In one possible implementation, when the sending module sends the encrypted data packet to the computing node through the encrypted transmission unit, it is specifically used to: perform difference detection on the historical data of the target subject and the encrypted data packet through the data difference unit to obtain difference data; and send the difference data to the computing node.

[0019] In one possible implementation, the decryption module decrypts the encrypted data packet in the computing node to obtain decrypted data, specifically by: verifying the identity of the data requester through the distributed identity authentication unit, tagging the authenticated data requester to obtain an authentication result; and decrypting the corresponding encrypted data packet based on a preset decryption strategy according to the authentication result to obtain the decrypted data.

[0020] In one possible implementation, before sending the encrypted data packet to the computing node via the encrypted transmission unit, the sending module is further configured to: dynamically authorize and verify the access policy of the data provider node through the smart contract unit to obtain the compliant data provider node, wherein the dynamic authorization verification includes comparing the access range of the request field of the data requester with the preset field of the data provider; at the compliant data provider node, encrypting the provided data based on a preset encryption method to obtain the encrypted data packet, wherein the preset encryption method includes attribute-based encryption and homomorphic encryption, wherein the attribute-based encryption is used to dynamically control data access permissions according to the attributes of the decryptor, and the homomorphic encryption is used to directly perform computation operations without decryption.

[0021] In one possible implementation, when the construction module constructs the first business model in the distributed data sharing system based on the fused feature dataset, it is specifically used to: construct an operations research model according to the objective function, the constraints, and the decision variables; and, based on the distributed data sharing system, call the fused feature dataset to solve the operations research model to obtain the first business model.

[0022] In one possible implementation, the operations research model includes at least one of the following: a linear programming model for maximizing expected returns under capital constraints; and a stochastic programming model for handling uncertainty constraints.

[0023] In one possible implementation, the method is further used to: upload the generation process of the fused feature dataset and the first business model to the blockchain supervision node through the evidence storage unit to form an immutable audit log.

[0024] Thirdly, embodiments of this disclosure provide an electronic device, including: a processor, and a memory communicatively connected to the processor;

[0025] The memory stores computer-executed instructions;

[0026] The processor executes computer execution instructions stored in the memory to implement the method as described in any one of claims 1 to 7.

[0027] Fourthly, embodiments of this disclosure provide a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the first aspect and / or various possible implementations of the first aspect.

[0028] Fifthly, embodiments of this disclosure provide a computer program product, including a computer program that, when executed by a processor, implements the first aspect and / or various possible implementations of the first aspect.

[0029] This application provides a model creation method, apparatus, device, storage medium, and program product. The method involves sending encrypted data packets to a computing node via an encrypted transmission unit; decrypting the encrypted data packets on the computing node to obtain decrypted data; allocating isolated memory to extract features from the decrypted data to obtain a fused feature dataset; constructing a first business model based on the fused feature dataset in a distributed data sharing system; the data provider encrypts local data and transmits it to the computing node, then decrypts it, confirming the same entity identifier recorded in different data sources without disclosing data from any party; subsequently, data cleaning (e.g., filling in missing values) and feature extraction (e.g., normalization) are performed in isolated memory to generate a standardized fused feature dataset; finally, a globally optimal decision based on multi-source data is achieved through an operations research model. This resolves the conflict between cross-institutional credit data fusion and privacy protection. Attached Figure Description

[0030] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0031] Figure 1 A schematic diagram illustrating the creation of a financial lending platform based on a cross-agency credit data system, as provided in this embodiment of the disclosure;

[0032] Figure 2 A flowchart illustrating a model creation method provided in this embodiment of the disclosure. Figure One ;

[0033] Figure 3 A flowchart illustrating a model creation method provided in this embodiment of the disclosure. Figure Two ;

[0034] Figure 4 A flowchart illustrating a model creation method provided in this embodiment of the disclosure. Figure Three ;

[0035] Figure 5 A system architecture diagram of a model creation method provided in this disclosure embodiment;

[0036] Figure 6 A schematic diagram of the structure of a model creation apparatus provided in this embodiment of the disclosure. Figure One ;

[0037] Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this disclosure.

[0038] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0039] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0040] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, storage, use, processing, transmission, provision, disclosure, and application of the relevant data all comply with the relevant laws, regulations, and standards of the relevant countries and regions, have taken necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation access points for users to choose to authorize or refuse.

[0041] Furthermore, the technical solution involved in this application, which involves big data analysis of user information (including but not limited to personal biometrics, identity data, consumption data, asset data, electronic terminal operation data, etc.) and the use of artificial intelligence technology for automated decision-making, and makes decisions that have a significant impact on personal rights based on the results of automated decision-making, provides users with corresponding operation entry points for users to choose to agree to or reject the results of automated decision-making; if the user chooses to reject, the process will proceed to the expert decision-making process.

[0042] It should be noted that the model creation method, apparatus, device, storage medium and program product provided in this application can be used in the field of financial technology, or in any field other than financial technology. The application field is not limited in this application.

[0043] The specific application scenarios of this disclosure are as follows. Figure 1 This is a schematic diagram illustrating the creation of a financial lending platform based on a cross-institutional credit data system, as provided in this embodiment. First, the system adopts a distributed architecture, including multiple data provider nodes (such as data node 1, data node 2, data node 3, etc.), computing nodes (deploying a trusted execution environment cluster), and blockchain monitoring nodes. Data providers encrypt and upload their local credit data to the computing nodes. The computing nodes, within the trusted execution environment, complete data decryption, fusion, and operations research model solving, ultimately returning the decision result to the requesting party. This result is then stored on the blockchain, meaning the local credit data and the process of generating the first business model are uploaded to the blockchain monitoring node. For example, in the approval of micro and small enterprise loans, banks do not need to obtain the enterprise's original tax, social security, and transaction records; they can allocate the globally optimal loan amount based on multi-source fusion characteristics, ensuring that the data provider nodes can complete the entire process traceability through the blockchain monitoring node.

[0044] This disclosure provides a model creation method that aims to solve the above-mentioned technical problems in the prior art.

[0045] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0046] Figure 2 A flowchart illustrating a model creation method provided in this embodiment of the disclosure. Figure One For example, such as Figure 2 As shown, the method includes:

[0047] Step S101: Send the encrypted data packet to the computing node through the encrypted transmission unit.

[0048] Step S102: Decrypt the encrypted data packet in the computing node to obtain the decrypted data.

[0049] Step S103: Allocate isolated memory to extract features from the decrypted data to obtain a fused feature dataset.

[0050] Step S104: Based on the fused feature dataset, construct the first business model in the distributed data sharing system. The first business model is implemented based on the operations research model.

[0051] For example, the compliant data provider node refers to the institutional service node that stores encrypted credit data; the encrypted transmission unit refers to the system's built-in data transmission component, responsible for the secure routing and network transmission of encrypted data packets, ensuring that the transmission process is not stolen or tampered with; the computing node refers to the server cluster equipped with a trusted execution environment, responsible for data decryption, feature extraction, operations research model construction and solution; isolated memory refers to an independent secure memory area within the computing node, physically / logically isolated from the system's main memory, where plaintext data is processed only to prevent data leakage; the fused feature dataset refers to a standardized set of features formed after cleaning, aligning, and extracting multi-source credit data, such as combinations of features like the average corporate tax payment, social security payment stability, and historical overdue number; in one possible implementation, firstly, the encrypted transmission unit adopts a dedicated line encrypted transmission mode, establishing a dedicated intranet line between the data provider and the computing node, coupled with SSL transmission encryption, the encrypted transmission unit receives encrypted data packets output by each compliant data provider node, routes and forwards the data packets to the designated computing node according to preset transmission rules, and uses encrypted channels for transmission throughout, preventing external devices from intercepting valid data. Next, the compute node invokes its built-in decryption module to decrypt the encrypted data packet using the corresponding decryption key. The decrypted plaintext data is temporarily stored in the system cache for further processing. Then, the compute node actively allocates an independent, isolated memory region and imports the decrypted data into this region. Data cleaning, entity alignment, feature filtering, and normalization are then performed sequentially, integrating multi-source credit data to generate a unified fused feature dataset. After processing, plaintext data cannot be read outside the isolated memory region. Finally, the compute node reads the fused feature dataset, invokes the operations research engine, and builds the first business model for credit based on business rules, completing model initialization and parameter configuration. In another possible implementation, the encrypted transmission unit uses a VPN encrypted transmission mode, establishing an encrypted virtual channel over the public network for encrypted data transmission. Encrypted data packets are sent through the encrypted transmission unit, the compute node decrypts the data, allocates isolated memory for feature extraction, and finally constructs the first business model based on operations research using the fused feature dataset.

[0052] The above steps rely on encrypted transmission and isolated memory to break down data silos. At the same time, they strictly protect data privacy through encryption and isolation, solving the problems of single data source and high risk of privacy leakage in traditional credit models, and effectively improving the comprehensiveness and reliability of financial decision-making models.

[0053] Furthermore, in one possible implementation, step S101 is specifically implemented as follows:

[0054] Step S101-1: Perform difference detection on the historical data of the target subject and the encrypted data packet through the data difference unit to obtain the difference data;

[0055] Step S101-2: Send the difference data to the computing node.

[0056] For example, the data difference unit is a built-in data comparison component of the system, used to compare historical data of the same target subject with the latest encrypted data packet, and identify the incremental and changed parts of the data. The difference data refers to the incremental data that has been added, modified, or changed after comparing the historical data with the latest data packet, such as the addition of tax records or the addition of overdue information for enterprises. In one possible implementation, firstly, the data difference unit retrieves the target subject's historical credit data stored locally, and uses a full-field line-by-line comparison algorithm to compare all contents of the historical data with the new data packet, and performs hash comparison and content verification with the currently received encrypted data packet field by field and item by item, automatically identifying the changed content, filtering out the difference data, and directly filtering out the data without changes. Afterwards, the encrypted transmission unit only packages and encrypts the filtered difference data and forwards it to the computing node, and the complete historical data is not transmitted again. In another possible implementation, firstly, the data difference unit uses a hash digest comparison algorithm to generate a hash value for the entire data, and performs a partial comparison if the hashes are inconsistent, and then sends the difference data to the computing node.

[0057] The above steps add a data differential unit to the basic transmission process. The data differential unit identifies incremental data and transmits only the different content. The differential detection filters out duplicate data and transmits only incremental data, thereby reducing the amount of data transmitted, reducing network bandwidth usage and data transmission latency.

[0058] Furthermore, in one possible implementation, step S102 is specifically implemented as follows:

[0059] Step S102-1: Verify the identity of the data requester through the distributed identity authentication unit, tag the data requester that passes the authentication, and obtain the authentication result.

[0060] Step S102-2: Based on the authentication result, decrypt the corresponding encrypted data packet according to the preset decryption strategy to obtain the decrypted data.

[0061] For example, the distributed identity authentication unit refers to a blockchain-based identity verification component used to verify the legitimacy of the data requester's identity; the data requester refers to the institution initiating the model creation and data usage request, which in this scenario refers to financial entities such as banks and credit institutions; the authentication result refers to the identification information after identity verification is completed. If authentication is successful, a legitimate identity tag is added; if it fails, the data is intercepted; the preset decryption policy refers to the pre-configured decryption rules, which are bound to the requester's identity and data usage permissions, with different identities corresponding to different decryption ranges; in one possible implementation, firstly, the distributed identity authentication unit reads the data requester's digital identity credential and compares it with the legitimate identity database stored on the chain. Before the comparison, the identity is statically bound to the decryption policy in advance, and permanent decryption permissions are configured for fixed financial institutions in advance. If the identity is legitimate, the request is tagged as authenticated; if the identity is abnormal, the process is terminated directly and an error is returned. Afterwards, the authentication tag is read by the computing node, matched with the preset decryption policy corresponding to the requester, and the encrypted data packet is decrypted in a targeted manner according to the permission range; sensitive fields outside the permission remain encrypted and are not parsed. Another possible implementation involves dynamically configuring identity and decryption strategies. Decryption permissions are temporarily assigned based on a single task and automatically reclaimed when the task ends. First, the distributed identity authentication unit completes identity verification and tagging. Then, the data is decrypted based on the authentication result and the preset decryption strategy.

[0062] The above steps add a distributed identity authentication and policy control process before data decryption. First, the identity of the requester is verified, and then targeted decryption is performed according to permissions. This realizes the verification of the requester's identity and fine-grained decryption permission control, preventing unauthorized entities from obtaining data.

[0063] Figure 3 A flowchart illustrating a model creation method provided in this embodiment of the disclosure. Figure Two For example, such as Figure 3 As shown, the method includes:

[0064] Step S201: Dynamically authorize and verify the access policy of the data provider node through the smart contract unit to obtain compliant data provider nodes.

[0065] Step S202: At the compliant data provider node, the provided data is encrypted using a preset encryption method to obtain an encrypted data packet.

[0066] Step S203: Send the encrypted data packet to the computing node through the encrypted transmission unit.

[0067] Step S204: Decrypt the encrypted data packet in the computing node to obtain the decrypted data.

[0068] Step S205: Allocate isolated memory to extract features from the decrypted data to obtain a fused feature dataset.

[0069] Step S206: Based on the fused feature dataset, construct the first business model in the distributed data sharing system. The first business model is implemented based on the operations research model.

[0070] Steps S203 to S206 have been explained in detail above and will not be repeated here, only briefly described.

[0071] For example, a blockchain regulatory node refers to a regulatory service node equipped with a consortium blockchain, which stores access policies, authorization records, and encrypted logs; a smart contract unit refers to an automated execution script deployed on the blockchain for dynamic authorization verification; an access policy refers to the data usage rules preset by the data provider, including accessible institutions, queryable fields, and usage scenarios, such as "only banks are allowed to query the tax amount of micro and small enterprises, and querying shareholder information is prohibited"; attribute-based encryption refers to an encryption algorithm that allocates decryption permissions based on the attributes of the access subject, such as only credit institutions can decrypt credit data; homomorphic encryption refers to an encryption algorithm that supports direct computation in the encrypted state, allowing simple statistical and calculation operations to be completed before decryption; in one possible implementation, the smart contract unit automatically retrieves the access policy preset by the data provider, compares the request fields and usage scenarios submitted by the data requester with the policy one by one, and if the verification passes, the node is determined to be a compliant data provider node, and if the verification fails, the request is rejected. Subsequently, the compliant data provider node selects an encryption algorithm based on the business scenario, using attribute-based encryption or homomorphic encryption to encrypt the original credit data, generating a standardized encrypted data packet. The encryption transmission unit then sends the encrypted data packet securely to the computing node. The computing node decrypts the data and, combined with the identity authentication result, completes the data decryption. Next, data processing and feature fusion are performed in secure, isolated memory, and features are extracted to obtain a fused feature dataset. Finally, the credit model (i.e., the first business model) is built based on the fused feature dataset. In another possible implementation, the smart contract unit performs dynamic authorization verification. The compliant node uses a homomorphic encryption algorithm to generate an encrypted data packet. The encryption transmission unit then sends the encrypted data packet to the computing node. The computing node decrypts the data, extracts features from isolated memory to obtain a fused feature dataset, and finally constructs the first business model based on operations research.

[0072] The above steps supplement the pre-authorization verification and data encryption process. Dynamic authorization is achieved by relying on blockchain smart contracts, and data encryption is completed by using two encryption algorithms. Then, the transmission, decryption, feature extraction, and model building processes are executed in sequence. Authorization verification is completed before data encryption and transmission, standardizing the data encryption method and further strengthening the security of the original data.

[0073] Furthermore, in one possible implementation, step S206 is specifically implemented as follows:

[0074] Step S206-1: Construct an operations research model based on the objective function, constraints, and decision variables;

[0075] Step S206-2: Based on the distributed data sharing system, call the fused feature dataset to solve the operations research model and obtain the first business model.

[0076] For example, the objective function refers to the optimization objective of the operations research model, the constraints refer to the rules limiting the model's operation, and the decision variables refer to the adjustable variables in the model. In one possible implementation, a linear programming model is used. First, the preset objective function, constraints, and decision variables are read into the computation node, and the model framework is built according to operations research specifications. The fused feature dataset is then bound as the model's input data source. Next, the operations research solver is called to iteratively calculate the model, achieving parameter convergence, and finally generating a first-stage credit business model that can be directly used. In another possible implementation, a stochastic programming model is used to construct the operations research model. Then, the dataset is called to solve for the first-stage business model.

[0077] The above steps break down the construction and solution process of operations research models, distinguish between two applicable models: linear programming and stochastic programming, and rely on multi-dimensional credit data to enable operations research models to break through the limitations of single data, achieve globally optimal decision-making, and improve the accuracy of credit allocation and risk assessment.

[0078] Furthermore, in one possible implementation, the operations research model includes at least one of the following:

[0079] Linear programming models are used to maximize expected returns under capital constraints.

[0080] Stochastic programming models are used to handle uncertain constraints.

[0081] For example, a linear programming model refers to a mathematical programming model where both the objective function and constraints are linear functions of the decision variables. It is suitable for conventional credit allocation scenarios where the market environment, business operations are stable, and constraints are fixed. A stochastic programming model refers to an optimization decision-making model where the objective function or constraints include random variables. It is suitable for risk pricing and dynamic credit granting scenarios where market fluctuations are large and business operations are uncertain.

[0082] Figure 4 A flowchart illustrating a model creation method provided in this embodiment of the disclosure. Figure Three For example, such as Figure 4 As shown, the method includes:

[0083] Step S301: Send the encrypted data packet to the computing node through the encrypted transmission unit.

[0084] Step S302: Decrypt the encrypted data packet in the computing node to obtain the decrypted data.

[0085] Step S303: Allocate isolated memory to extract features from the decrypted data to obtain a fused feature dataset.

[0086] Step S304: Based on the fused feature dataset, construct the first business model in the distributed data sharing system. The first business model is implemented based on the operations research model.

[0087] Step S305: Through the evidence storage unit, the generation process of the fused feature dataset and the first business model are uploaded to the blockchain supervision node to form an immutable audit log.

[0088] Steps S301 to S304 have been explained in the previous text and will not be repeated here. Only step S305 will be introduced here.

[0089] For example, in one possible implementation, the evidence storage unit automatically collects information such as the logs generated from the fusion feature dataset, parameters of the first business model, and operation records. This includes process logs, complete model parameters, and detailed data, all packaged and uploaded to the blockchain monitoring node to complete the full on-chain information upload and generate a permanent, tamper-proof audit log. In another possible implementation, hash digests are used for on-chain upload, only the hash values ​​of the data and model are uploaded, while the audit log is stored locally.

[0090] The above steps generate audit logs by uploading all process information to the blockchain through the evidence storage unit. This not only enables data security modeling and global decision-making, but also supplements the full-process traceability capability, thereby improving the system's credibility and regulatory compatibility.

[0091] Figure 5 The system architecture diagram of a model creation method provided in this embodiment shows the overall layered structure, physical nodes and logical functional units of the system. At the top of each part is the business application layer, which includes the corresponding data requesters, mainly financial entities such as banks, credit institutions and financial risk control platforms. It is responsible for initiating data access requests and business model creation requests, and is the final business user of this solution.

[0092] The second layer consists of blockchain regulatory nodes, which serve as the core regulatory, authorization, and evidence storage nodes of the system. They integrate four logical units: 1. A distributed identity authentication unit (based on blockchain to achieve decentralized identity verification, validating the legality of the data requester and data provider); 2. A smart contract unit (automatically executes data access policy verification, comparing requested fields with preset access ranges to complete dynamic authorization); 3. A blockchain ledger (permanently stores identity information, access policies, and authorization records; the data is tamper-proof); and 4. An evidence storage unit (collects information throughout the entire data processing and model creation process, generates audit logs, and uploads them to the blockchain).

[0093] The third layer consists of compliant data provider nodes, including distributed data source nodes. These nodes include institutions holding credit data such as tax bureaus, human resources and social security bureaus, the central bank's credit reference center, and third-party payment platforms. These serve as the system's original data source and integrate three functional units: 1. Local data source (stores original credit data for micro and small enterprises, such as tax payments, social security contributions, credit records, and transaction details); 2. Data differentiation unit (compares historical data with the latest data, identifies incremental differences, and reduces transmission overhead); 3. Encryption module (supports both attribute-based encryption and homomorphic encryption, encrypting the original data locally to ensure that the data is encrypted upon leaving the node).

[0094] The fourth layer is the encrypted transmission layer, which consists of encrypted transmission units. It uses encrypted transmission methods such as dedicated lines, VPNs, and SSL to securely route and transmit encrypted data packets / differential data, preventing data from being stolen or tampered with in the network link.

[0095] The final layer consists of computing nodes, which are the core computing nodes of the system. They are equipped with a trusted execution environment and serve as the core carrier for model creation. They integrate four functional modules: 1. Decryption module (combining identity authentication results with preset decryption strategies to perform targeted decryption of encrypted data); 2. Isolated memory (a secure memory area isolated by hardware and software, where plaintext data is processed only to prevent data leakage); 3. Feature extraction module (completing data cleaning, entity alignment, and feature normalization to generate a fused feature dataset); and 4. Operations research modeling engine (built-in operations research models such as linear programming and stochastic programming, which are used to build and solve the first business model (credit limit and risk assessment model) based on the fused feature dataset).

[0096] Figure 6 A schematic diagram of the structure of a model creation apparatus provided in this embodiment of the disclosure. Figure One ,like Figure 6 As shown, the model creation apparatus 40 provided in this embodiment includes:

[0097] Sending module 401 is used to send encrypted data packets to computing nodes through the encrypted transmission unit;

[0098] The decryption module 402 is used to decrypt encrypted data packets in the computing node to obtain decrypted data;

[0099] Extraction module 403 is used to allocate isolated memory to extract features from the decrypted data and obtain a fused feature dataset;

[0100] Module 404 is used to build a first business model in a distributed data sharing system based on a fused feature dataset. The first business model is implemented based on an operations research model.

[0101] In one possible implementation, when the sending module 401 sends the encrypted data packet to the computing node through the encrypted transmission unit, it is specifically used to: perform difference detection on the historical data of the target subject and the encrypted data packet through the data difference unit to obtain the difference data; and send the difference data to the computing node.

[0102] In one possible implementation, the decryption module 402 decrypts the encrypted data packet in the computing node to obtain decrypted data. Specifically, it performs the following steps: authenticates the data requester through a distributed identity authentication unit, tags the authenticated data requester, and obtains the authentication result; based on the authentication result, it decrypts the corresponding encrypted data packet according to a preset decryption strategy to obtain decrypted data.

[0103] In one possible implementation, before sending the encrypted data packet to the computing node via the encrypted transmission unit, the sending module 401 is further configured to: dynamically authorize and verify the access policy of the data provider node through the smart contract unit to obtain a compliant data provider node; the dynamic authorization verification includes comparing the access range of the request field of the data requester with the preset field of the data provider; at the compliant data provider node, encrypt the provided data based on a preset encryption method to obtain an encrypted data packet; the preset encryption method includes attribute-based encryption and homomorphic encryption; attribute-based encryption is used to dynamically control data access permissions according to the attributes of the decryptor; and homomorphic encryption is used to directly perform computation operations without decryption.

[0104] In one possible implementation, when the construction module 404 constructs the first business model in the distributed data sharing system based on the fused feature dataset, it is specifically used to: construct an operations research model according to the objective function, constraints, and decision variables; and, based on the distributed data sharing system, call the fused feature dataset to solve the operations research model to obtain the first business model.

[0105] In one possible implementation, the operations research model includes at least one of the following: a linear programming model for maximizing expected returns under capital constraints; and a stochastic programming model for handling uncertainty constraints.

[0106] In one possible implementation, the method is further used to: upload the generation process of the fused feature dataset and the first business model to the blockchain supervision node through the evidence storage unit to form an immutable audit log.

[0107] Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this disclosure. Figure 7 As shown, the electronic device 50 provided in this embodiment includes at least one processor 501 and a memory 502. Optionally, the device 50 further includes a communication component 503. The processor 501, memory 502, and communication component 503 are connected via a bus 504.

[0108] In a specific implementation, at least one processor 501 executes computer execution instructions stored in memory 502, causing at least one processor 501 to perform the above-described method.

[0109] The specific implementation process of processor 501 can be found in the above method embodiments, and its implementation principle and technical effect are similar. It will not be repeated here.

[0110] In the above embodiments, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.

[0111] The memory may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage device.

[0112] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.

[0113] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method.

[0114] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the above-described method.

[0115] The aforementioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.

[0116] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an Application Specific Integrated Circuit (ASIC). Alternatively, the processor and the readable storage medium can exist as discrete components in the device.

[0117] The division of units is merely a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0118] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0119] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0120] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0121] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0122] Finally, it should be noted that other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein, and is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of the invention is limited only by the appended claims.

[0123] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily essential to this application.

[0124] It should be further noted that although the steps in the flowchart are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowchart may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0125] It should be understood that the above-described device embodiments are merely illustrative, and the device of this application can also be implemented in other ways. For example, the division of units / modules in the above embodiments is only a logical functional division, and there may be other division methods in actual implementation. For example, multiple units, modules, or components may be combined, or integrated into another system, or some features may be ignored or not executed.

[0126] Furthermore, unless otherwise specified, the functional units / modules in the various embodiments of this application can be integrated into one unit / module, or each unit / module can exist physically separately, or two or more units / modules can be integrated together. The integrated units / modules described above can be implemented in hardware or as software program modules.

[0127] When integrated units / modules are implemented in hardware, the hardware can be digital circuits, analog circuits, etc. The physical implementation of the hardware structure includes, but is not limited to, transistors, memristors, etc. Unless otherwise specified, the processor can be any suitable hardware processor, such as a CPU, GPU, FPGA, DSP, and ASIC, etc. Unless otherwise specified, the storage unit can be any suitable magnetic or magneto-optical storage medium, such as Resistive Random Access Memory (RRAM), Dynamic Random Access Memory (DRAM), Static Random Access Memory (SRAM), Enhanced Dynamic Random Access Memory (EDRAM), High-Bandwidth Memory (HBM), Hybrid Memory Cube (HMC), etc.

[0128] If the integrated unit / module is implemented as a software program module and sold or used as an independent product, it can be stored in a computer-readable storage device (CMD). Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a memory and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned memory includes various media capable of storing program code, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard drive, magnetic disk, or optical disk.

[0129] In the above embodiments, the descriptions of each embodiment have their own emphasis. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments. The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as the combination of these technical features does not contradict each other, it should be considered within the scope of this specification.

[0130] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this application are indicated by the following claims.

[0131] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this application is limited only by the appended claims.

[0132] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily essential to this application.

[0133] It should be further noted that although the steps in the flowchart are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowchart may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0134] It should be understood that the above-described device embodiments are merely illustrative, and the device of this application can also be implemented in other ways. For example, the division of units / modules in the above embodiments is only a logical functional division, and there may be other division methods in actual implementation. For example, multiple units, modules, or components may be combined, or integrated into another system, or some features may be ignored or not executed.

[0135] Furthermore, unless otherwise specified, the functional units / modules in the various embodiments of this application can be integrated into one unit / module, or each unit / module can exist physically separately, or two or more units / modules can be integrated together. The integrated units / modules described above can be implemented in hardware or as software program modules.

[0136] When integrated units / modules are implemented in hardware, the hardware can be digital circuits, analog circuits, etc. The physical implementation of the hardware structure includes, but is not limited to, transistors, memristors, etc. Unless otherwise specified, the processor can be any suitable hardware processor, such as a CPU, GPU, FPGA, DSP, and ASIC, etc. Unless otherwise specified, the storage unit can be any suitable magnetic or magneto-optical storage medium, such as Resistive Random Access Memory (RRAM), Dynamic Random Access Memory (DRAM), Static Random Access Memory (SRAM), Enhanced Dynamic Random Access Memory (EDRAM), High-Bandwidth Memory (HBM), Hybrid Memory Cube (HMC), etc.

[0137] If the integrated unit / module is implemented as a software program module and sold or used as an independent product, it can be stored in a computer-readable storage device (CMD). Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a memory and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned memory includes various media capable of storing program code, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard drive, magnetic disk, or optical disk.

[0138] In the above embodiments, the descriptions of each embodiment have their own emphasis. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments. The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as the combination of these technical features does not contradict each other, it should be considered within the scope of this specification.

[0139] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this application are indicated by the following claims.

[0140] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this application is limited only by the appended claims.

Claims

1. A model creation method applied to a distributed data sharing system, the distributed data sharing system comprising compliant data provider nodes, computing nodes, and encrypted transmission units, wherein the compliant data provider nodes store encrypted data packets, characterized in that, The method includes: The encrypted data packet is sent to the computing node through the encrypted transmission unit; The encrypted data packet is decrypted in the computing node to obtain decrypted data; An isolated memory is allocated to extract features from the decrypted data, resulting in a fused feature dataset; Based on the fused feature dataset, a first business model is constructed in the distributed data sharing system. The first business model is implemented based on an operations research model.

2. The method according to claim 1, wherein the distributed data sharing system further includes a data differential unit, characterized in that, Sending the encrypted data packet to the computing node via the encrypted transmission unit includes: The difference data is obtained by performing difference detection between the historical data of the target subject and the encrypted data packet through the data difference unit; The difference data is sent to the computing node.

3. The method according to claim 1, wherein the distributed data sharing system further includes a distributed identity authentication unit, characterized in that, The process of decrypting the encrypted data packet in the computing node to obtain decrypted data includes: The distributed identity authentication unit authenticates the data requester, tags the authenticated data requester, and obtains the authentication result. Based on the authentication result, the corresponding encrypted data packet is decrypted using a preset decryption strategy to obtain the decrypted data.

4. The method according to claim 1, wherein the distributed data sharing system further includes a blockchain monitoring node, the blockchain monitoring node being deployed with smart contract units, characterized in that, Before sending the encrypted data packet to the computing node via the encrypted transmission unit, the method further includes: The access policy of the data provider node is dynamically authorized and verified through the smart contract unit to obtain the compliant data provider node. The dynamic authorization verification includes comparing the access range of the request field of the data requester with the preset field of the data provider. At the compliant data provider node, the provided data is encrypted based on a preset encryption method to obtain the encrypted data packet. The preset encryption method includes attribute-based encryption and homomorphic encryption. The attribute-based encryption is used to dynamically control data access permissions according to the attributes of the decryptor, and the homomorphic encryption is used to directly perform calculation operations without decryption.

5. The method according to claim 1, wherein the computing node is pre-defined with an objective function, constraints, and decision variables, characterized in that, The construction of a first business model in the distributed data sharing system based on the fused feature dataset includes: Based on the objective function, the constraints, and the decision variables, construct an operations research model; Based on the distributed data sharing system, the fused feature dataset is invoked to solve the operations research model, thereby obtaining the first business model.

6. The method according to claim 5, characterized in that, The operations research model includes at least one of the following: Linear programming models are used to maximize expected returns under capital constraints. Stochastic programming models are used to handle uncertain constraints.

7. The method according to claim 1, wherein the distributed data sharing system further includes a blockchain monitoring node, and the blockchain monitoring node further includes an evidence storage unit, characterized in that, The method further includes: The evidence storage unit uploads the generation process of the fused feature dataset and the first business model to the blockchain supervision node to form an immutable audit log.

8. A model creation apparatus, comprising: The sending module is used to send the encrypted data packet to the computing node through the encrypted transmission unit; A decryption module is used to decrypt the encrypted data packet in the computing node to obtain decrypted data; The extraction module is used to allocate isolated memory to extract features from the decrypted data to obtain a fused feature dataset; A construction module is used to construct a first business model in the distributed data sharing system based on the fused feature dataset. The first business model is implemented based on an operations research model.

9. An electronic device, characterized in that, include: A processor, and a memory communicatively connected to the processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory to implement the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method as described in any one of claims 1 to 7.

11. A computer program product, characterized in that, Includes a computer program that, when executed by a processor, implements the method of any one of claims 1 to 7.

Images