Block chain and deep learning-based delayed clearing method, device and equipment, medium and program product

Abstract

The invention provides a delayed clearing method and device based on a block chain and deep learning, equipment, a medium and a program product, and relates to the technical field of block chains, and the method comprises the steps: receiving a transaction proposal sent by a client, verifying the transaction proposal through an endorsement node of the block chain, and generating signature verification information; sending the signature verification information to the client, and receiving merchant information of a plurality of merchants whose signature verification information satisfies a preset condition, wherein the merchant information is sent by the client; multiple pieces of merchant information are preprocessed and then input into a pre-trained neural network model, respective delayed clearing permissions of all merchants are output, and the number of hidden layers, used for processing the merchant information, of the neural network model and the number of neurons of each hidden layer are dynamically determined based on the number of variables contained in the merchant information; calling an intelligent contract to store archive information of merchants whose settlement delay authority is allowed to access; and in response to a delayed clearing request of the target merchant sent by the client, executing delayed clearing for the target merchant based on the archive information.

Description

Technical Field

[0001] This application relates to the field of blockchain technology, and can be specifically applied to the field of financial technology. More specifically, it relates to a delayed settlement method, device, equipment, medium and program product based on blockchain and deep learning. Background Art

[0002] Bank acquiring businesses are shifting resources from focusing on the number of acquiring merchants to the quality of those merchants. Upgrading acquiring systems to support features like deferred settlement will help improve customer loyalty for large-scale merchants. Existing acquiring technology relies on manual identification of merchants eligible for deferred settlement, resulting in low accuracy and efficiency, high workload, and data security risks. Summary of the Invention

[0003] In view of the above problems, the present application provides a delayed settlement method, device, equipment, medium and program product based on blockchain and deep learning to improve recognition accuracy, recognition efficiency and data security.

[0004] According to the first aspect of the present application, a delayed settlement method based on blockchain and deep learning is provided, which is applied to blockchain, and the method includes: receiving a transaction proposal sent by a client, verifying the transaction proposal through an endorsement node of the blockchain, and generating signature verification information, wherein the transaction proposal includes a smart contract identifier, a smart contract method, parameter information for calling the smart contract, and signature information of the client; sending signature verification information to the client, and receiving merchant information of multiple merchants sent by the client, wherein the multiple merchants are merchants whose signature verification information meets preset conditions; after pre-processing the merchant information of the multiple merchants, respectively inputting the pre-trained neural network model, and outputting the delayed settlement rights of each merchant, the number of hidden layers and the number of neurons in each hidden layer in the neural network model for processing merchant information are determined based on the number of variables contained in the merchant information; calling the smart contract to store the archive information of merchants whose delayed settlement rights are allowed to access; and responding to the deferred settlement request of the target merchant sent by the client, performing delayed settlement for the target merchant based on the pre-stored archive information.

[0005] According to an embodiment of the present application, preprocessing merchant information of multiple merchants includes: performing feature standardization on the merchant information of the multiple merchants, and converting the merchant information into digital data of a preset interval.

[0006] According to an embodiment of the present application, merchant information of multiple merchants is preprocessed and then respectively input into a pre-trained neural network model to output the delayed settlement authority of each merchant, including: performing nonlinear transformation and compression on the merchant information through the hidden layer of the neural network model to capture the nonlinear relationship between features in the merchant information and generate first abstract features of multiple dimensions; performing deep fusion and compression on the first abstract features of multiple dimensions through the hidden layer of the neural network model to generate second abstract features; determining the merchant's label based on the second abstract feature, and the label is used to indicate the merchant's delayed settlement authority.

[0007] According to an embodiment of the present application, the number of hidden layers is determined by rounding the logarithmic operation result of the number of variables included in the merchant information, and the number of neurons is determined based on the number of variables included in the merchant information and the number of hidden layers.

[0008] According to an embodiment of the present application, delayed settlement is performed for the target merchant based on pre-stored archival information, including: temporarily storing the merchant's consumption funds within the delayed settlement date in an internal account; responding to a return request initiated by a customer, calling a smart contract to deduct the relevant funds from the internal account and return them to the customer, and synchronously updating the fund information stored in the blockchain.

[0009] According to an embodiment of the present application, performing delayed liquidation for target merchants based on pre-stored archival information also includes: for merchants that exceed the liquidation date range and have any surplus, calling a smart contract to deduct relevant funds from an internal account for the merchant to credit; when the funds in the internal account exceed a preset threshold, storing the excess funds in a designated account and synchronously updating the fund information stored in the blockchain. According to an embodiment of the present application, the method also includes: responding to deferred liquidation requests from different target merchants sent by the client, assigning the deferred liquidation requests from different target merchants to different processing channels in the blockchain, and performing delayed liquidation for different target merchants based on the pre-stored archival information.

[0010] According to an embodiment of the present application, the method further includes inputting the updated blockchain data into a pre-trained neural network model, and updating the parameters of the pre-trained neural network model.

[0011] The second aspect of the present application provides a delayed settlement method based on blockchain and deep learning, which is applied to a client, and the method includes: receiving signature verification information sent by an endorsement node of the blockchain, the signature verification information is generated by the endorsement node based on a transaction proposal, and the transaction proposal includes a smart contract identifier, a smart contract method, parameter information for calling the smart contract, and the signature information of the client; entering the merchant information of the merchant whose signature verification information meets the preset conditions and sending it to the blockchain, so that the blockchain pre-processes the merchant information of multiple merchants and inputs them into a pre-trained neural network model respectively, and outputs the delayed settlement authority of each merchant. The number of hidden layers and the number of neurons in each hidden layer in the neural network model for processing merchant information are dynamically determined based on the number of variables contained in the merchant information; sending the archive information of the merchants whose delayed settlement authority is allowed to be accessed to the blockchain for storage; responding to the target merchant inputting a deferred settlement request, sending the deferred settlement request to the blockchain, so that the blockchain performs delayed settlement for the target merchant based on the pre-stored archive information.

[0012] According to an embodiment of the present application, the method further includes: constructing a visualization interface; responding to the merchant opening the visualization interface through the client's browser, entering the merchant's merchant information and profile information through the visualization interface, and displaying the results of delayed settlement to the merchant.

[0013] According to a third aspect of the present application, a blockchain-based and deep learning-based acquiring and delayed settlement device is provided, which is applied to a blockchain. The device includes: a first receiving module, configured to receive a transaction proposal sent by a client, verify the transaction proposal through an endorsement node of the blockchain, and generate signature verification information, wherein the transaction proposal includes a smart contract identifier, a smart contract method, parameter information for calling the smart contract, and the client's signature information; a sending and receiving module, configured to send the signature verification information to the client and receive merchant information of multiple merchants sent by the client, wherein the multiple merchants are merchants whose signature verification information meets preset conditions; a preprocessing module, configured to preprocess the merchant information of the multiple merchants and input the pretrained neural network model into the pretrained neural network model respectively, and output the delayed settlement authority of each merchant, wherein the number of hidden layers and the number of neurons in each hidden layer in the neural network model for processing the merchant information are dynamically determined based on the number of variables contained in the merchant information; a calling module, configured to call the smart contract to store profile information of merchants whose delayed settlement authority is allowed to access; and a first responding module, configured to respond to a deferred settlement request of a target merchant sent by the client, and perform delayed settlement for the target merchant based on the pre-stored profile information.

[0014] A fourth aspect of the present application provides a delayed settlement device for acquiring merchants based on blockchain and deep learning, which is applied to a client, and the device includes: a second receiving module, which is used to receive signature verification information sent by an endorsement node of the blockchain, the signature verification information being generated by the endorsement node based on a transaction proposal, the transaction proposal including a smart contract identifier, a smart contract method, parameter information for calling the smart contract, and signature information of the client; an input module, which is used to input merchant information of merchants whose signature verification information meets preset conditions and send it to the blockchain, so that the blockchain pre-processes the merchant information of multiple merchants and inputs them into a pre-trained neural network model respectively, and outputs the delayed settlement authority of each merchant. The number of hidden layers and the number of neurons in each hidden layer in the neural network model for processing merchant information are dynamically determined based on the number of variables contained in the merchant information; a first sending module, which is used to send the archive information of merchants whose delayed settlement authority is allowed to be accessed to the blockchain for storage; and a second sending module, which is used to respond to a target merchant inputting a delayed settlement request and send the delayed settlement request to the blockchain, so that the blockchain performs delayed settlement for the target merchant based on the pre-stored archive information.

[0015] The fifth aspect of the present application provides an electronic device, comprising: one or more processors; a memory for storing one or more computer programs, wherein the one or more processors execute the one or more computer programs to implement the steps of the above method.

[0016] The sixth aspect of the present application further provides a computer-readable storage medium having a computer program or instruction stored thereon, which implements the steps of the above method when the above computer program or instruction is executed by a processor.

[0017] The seventh aspect of the present application further provides a computer program product, comprising a computer program or instructions, which implement the steps of the above method when executed by a processor. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The above contents and other objects, features and advantages of the present application will become more apparent through the following description of the embodiments of the present application with reference to the accompanying drawings, in which:

[0019] Figure 1 Schematically illustrates an application scenario diagram of a delayed settlement method, apparatus, device, medium, and program product based on blockchain and deep learning according to an embodiment of the present application;

[0020] Figure 2 The flowchart of the delayed settlement method based on blockchain and deep learning according to one embodiment of the present application is schematically shown;

[0021] Figure 3Schematically illustrates a flow chart of an application client interacting with a blockchain to implement a delayed settlement transaction according to an embodiment of the present application;

[0022] Figure 4 Schematically shows a logical structure diagram of a modular component according to an embodiment of the present application;

[0023] Figure 5 Schematically illustrates a delayed settlement method based on blockchain and deep learning according to another embodiment of the present application;

[0024] Figure 6 The following schematically shows a block diagram of a delayed settlement device based on blockchain and deep learning according to an embodiment of the present application;

[0025] Figure 7 Schematically shows a structural block diagram of a delayed settlement device based on blockchain and deep learning according to another embodiment of the present application; and

[0026] Figure 8 A block diagram of an electronic device suitable for implementing a delayed settlement method based on blockchain and deep learning according to an embodiment of the present application is schematically shown. DETAILED DESCRIPTION

[0027] Hereinafter, embodiments of the present application will be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of the present application. In the detailed description below, for ease of explanation, many specific details are set forth to provide a comprehensive understanding of the embodiments of the present application. However, it is apparent that one or more embodiments may also be implemented without these specific details. In addition, in the following description, descriptions of known structures and technologies are omitted to avoid unnecessarily confusing the concepts of the present application.

[0028] The terms used herein are only for describing specific embodiments and are not intended to limit this application. The terms "comprise," "include," etc. used herein indicate the presence of the features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0029] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art unless otherwise defined. It should be noted that the terms used herein should be interpreted as having a meaning consistent with the context of this specification and should not be interpreted in an idealized or overly rigid manner.

[0030] When expressions such as "at least one of A, B, and C, etc." are used, they should generally be interpreted in accordance with the meaning commonly understood by those skilled in the art (for example, "a system having at least one of A, B, and C" should include but is not limited to a system having A alone, B alone, C alone, A and B, A and C, B and C, and / or A, B, C, etc.).

[0031] In the technical solution of this application, the user information involved (including but not limited to user personal information, user image information, user device information, such as location information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) are all information and data authorized by the user or fully authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of the relevant data comply with relevant laws, regulations and standards, take necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0032] In the scenario of using personal information for automated decision-making, the methods, devices, and systems provided in the embodiments of the present application all provide users with corresponding operation portals for users to choose to agree or reject the automated decision-making results; if the user chooses to reject, the expert decision-making process will be entered. The expression "automated decision-making" here refers to the activity of automatically analyzing and evaluating an individual's behavioral habits, interests and hobbies, or economic, health, credit status, etc. through computer programs and making decisions. The expression "expert decision-making" here refers to the activity of making decisions by people who specialize in a certain field, have specialized experience, knowledge and skills, and have reached a certain level of professionalism.

[0033] Before introducing the delayed settlement method based on blockchain and deep learning in this application, the key terms involved in this application are defined as follows:

[0034] Blockchain: A brand-new technology architecture that ensures data consistency within a distributed database, the blockchain, to achieve decentralization and immutability. A blockchain consists of a chain of blocks. Each block stores specific information, and they are linked together in chronological order. This chain is stored across all file servers. As long as at least one server in the system is functioning, the entire blockchain is secure. These servers, called nodes in the blockchain, provide storage space and computing power for the entire blockchain system. Modifying blockchain information requires the consent of half of the nodes and the modification of all nodes. Since these nodes are often controlled by different entities, tampering with blockchain information is extremely difficult.

[0035] The core technologies of blockchain include peer-to-peer transmission, encryption algorithms, and consensus mechanisms.

[0036] Peer-to-peer transmission refers to the use of P2P network protocols to achieve end-to-end data transmission. Exchanges (such as transfers and transactions) within the blockchain do not require confirmation and authorization from a third-party guarantee agency.

[0037] Encryption algorithms are used to encrypt information to improve data security.

[0038] Consensus refers to an algorithm that reaches consensus on the order of events within a certain timeframe. It is the foundation and core of blockchain technology. On a blockchain, every node maintains a copy of the ledger of all transactions recorded on the chain. Each new transaction is received at a different time by each node. To prevent corrupted information from being written into the block, all nodes must confirm the received message.

[0039] Distributed ledgers offer confidentiality, scalability, flexibility, and extensibility. They feature pluggable, independent, and functionally distinct modules, making them suitable for acquiring and deferred clearing businesses. The overall architecture of distributed ledgers is divided into the network layer, the core layer, and the interface layer. The core layer comprises member services, blockchain services, and chaincode services. The interface layer accesses identity, ledger, transaction, and smart contract information through interfaces and events. The network layer implements the P2P network and ensures the consistency of distributed blockchain storage.

[0040] Decentralization generally refers to running on a distributed network, where participants' information is securely protected (and potentially anonymous). Decentralized systems offer three advantages: fault tolerance: decentralized systems are less likely to cease functioning due to a single, unexpected local failure; attack resistance: the cost of attacking a decentralized system is higher than that of a centralized system; and collusion resistance: participants in a decentralized system are less likely to collude with one another.

[0041] Asymmetric encryption: Asymmetric encryption is a method for maintaining key confidentiality. It requires two keys: a public key (referred to as the public key) and a private key (referred to as the private key). The public and private keys form a pair. If data is encrypted with the public key, it can only be decrypted with the corresponding private key. Because encryption and decryption use two different keys, this algorithm is called an asymmetric encryption algorithm.

[0042] Member Services: This service provides member services, including registration, enrollment, and certificate application. Considering the security, privacy, regulatory oversight, auditing, and performance requirements of commercial applications, nodes and members must obtain certificates before they can join the blockchain network, significantly improving the security of the delayed settlement function.

[0043] Blockchain service: Responsible for the calculation and storage of distributed ledgers, sorting services between nodes, endorsement verification management, and the implementation of ledger storage methods. It is the core component of the blockchain and provides underlying support for the main functions of the blockchain.

[0044] Chaincode Service: Chaincode, also known as a smart contract, is a standard-based piece of code that implements specific business logic. Chaincode is decoupled from the underlying ledger and runs in a separate container, allowing updates to the chaincode to persist without affecting existing data. The chaincode in this application's delayed clearing blockchain is written in Java and runs within a container. After installation and instantiation, the chaincode connects to nodes within the same channel.

[0045] Interfaces and Events: Provide APIs for third-party applications to call, facilitating secondary development, and already supports Java language interfaces. Chaincodes can be installed and tested through SDK or command-line tools (CTL), and transaction status and data can be queried. Events can also be used to monitor events occurring in the blockchain network, facilitating the calling and processing of third-party application systems.

[0046] Network protocol: Implement P2P network transmission using gPRC and Gossip protocols.

[0047] A smart contract is a computer protocol designed to communicate, verify, or enforce contracts in an information-based manner. Smart contracts allow for trusted, traceable, and irreversible transactions without a third party. They aim to provide superior security to traditional contracts and reduce other transaction costs associated with them.

[0048] Endorsing Node: The node that undertakes the endorsement task in the blockchain is called an endorsing node. The endorsing node must prove its legitimacy by signing the expiration information of the valid certificate.

[0049] Anchor Node: In a channel, an anchor node can be discovered and communicated with by any node in the channel. Therefore, every organization that joins the channel has at least one anchor node. Nodes in one organization can discover all nodes in other organizations in the channel by searching for anchor nodes.

[0050] Leader node: Connects to the sorting service node as a representative within the organization, and distributes the batch blocks received from the sorting service node to other nodes within the organization.

[0051] Chaincode: Fabric's smart contract is called chaincode, which is divided into system chaincode and user chaincode. System chaincode is used to implement system-level functions, and user chaincode is used to implement user application functions.

[0052] Consortium blockchains: Consortium blockchains are open blockchain systems that are open to specific organizations. Obviously, since only certain nodes are allowed to connect to the blockchain system, this permission mechanism creates a potential for centralization. However, if only specific users have access to the private keys for specific transactions, the consortium chain can be further decentralized.

[0053] Consensus algorithm: A set of processes and rules created to implement a distributed consensus protocol. When nodes distributed across different domains negotiate and interact according to this set of rules, they can ultimately reach a consensus on a particular issue or issues, thereby achieving consistency across the distributed nodes.

[0054] X.509: X.509 is the format standard for public key certificates in cryptography. An X.509 certificate contains a public key, identity information (such as the network host name, organization name, or individual name), and signature information (signature of the certificate issuing authority or self-signature).

[0055] Authentication: Authentication, also known as verification or authorization, refers to the process of confirming a user's identity through certain means. There are many authentication methods, but this application uses a public-key encryption algorithm.

[0056] MSP: An MSP is a modular component abstracted from a distributed ledger that defines identity verification, authentication, and enforcement rules for network access. More specifically, an MSP is the component of a distributed ledger that manages and authenticates the identities of network members. This includes managing user IDs, verifying nodes seeking to join the network, and providing credentials for transactions initiated by clients.

[0057] CA: CA is the international term for a certification authority, an organization that issues, manages, and revokes digital certificates. A CA's role is to verify the legitimacy of the certificate holder's identity and digitally sign the certificate to prevent forgery or tampering.

[0058] Deep learning: Deep learning is used to build advanced neural networks. By establishing multiple hidden layers in the neural network, it can process complex nonlinear representations in the data and has high accuracy for complex multi-dimensional nonlinear scenarios.

[0059] Delayed Settlement: As acquiring businesses evolve from focusing on the number of acquiring merchants to the quality of those merchants, the need for new, specialized acquiring services has become a top priority. For example, in traditional processes, merchant returns do not support netting returns against a transitional internal account. The newly launched delayed settlement feature allows orders that have been cleared but not yet posted to the account to be held in a newly created delayed-posting internal account. Funds in this new internal account can then be used to net returns when processing refunds.

[0060] Input Layer: The architecture consists of a set of inputs, a processor, and a set of outputs, where the inputs constitute the input layer.

[0061] Output layer: The architecture consists of a set of inputs, a processor, and a set of outputs, where the outputs constitute the output layer.

[0062] Hidden layer: The architecture consists of a set of inputs, a processor, and a set of outputs, where the intermediate layer that performs the processing is called the hidden layer.

[0063] Weights: Weights are the most important factor in transforming inputs into outputs. This is similar to the slope in linear regression, where weights are multiplied by the inputs and summed together to form the output. Weights are numerical parameters that determine the influence of each neuron on another neuron.

[0064] Forward propagation: The process from the input layer to the hidden layer and then to the output layer is called forward propagation. At each layer, sum(input * weight) + bias is applied, and the activation function value is propagated to the next layer. The next layer can be a hidden layer or the output layer. Neural networks are constructed using a large number of hidden layers to produce deep neural networks.

[0065] Backward Propagation: The backpropagation process uses the partial derivatives of the activation function of each neuron to identify the slope (or gradient) in each direction going into the weights.

[0066] Learning rate: The gradient indicates how much a change in weights reduces or improves the error. The amount by which backpropagation continuously changes weights until the error is minimized is called the learning rate. The learning rate is a scalar parameter, similar to the step size in numerical integration, used to set the rate of adjustment to achieve faster error reduction. The learning rate is used in backpropagation when adjusting weights and biases.

[0067] Confusion matrix: A confusion matrix is ​​a table used to evaluate model performance in classification problems. It shows how the model classifies samples. The rows of the confusion matrix represent the actual categories and the columns represent the predicted categories.

[0068] Figure 1 The following schematically illustrates an application scenario diagram of a delayed settlement method, apparatus, device, medium, and program product based on blockchain and deep learning according to an embodiment of the present application.

[0069] like Figure 1 As shown, the application scenario 100 according to this embodiment may include a first terminal device 101, a second terminal device 102, a third terminal device 103, a network 104, and a server 105. The network 104 is used as a medium for providing a communication link between the first terminal device 101, the second terminal device 102, the third terminal device 103, and the server 105. The network 104 may include various connection types, such as wired or wireless communication links or optical fiber cables.

[0070] A user may use a first terminal device 101, a second terminal device 102, or a third terminal device 103 to interact with a server 105 via a network 104 to receive or send messages, etc. Various communication client applications may be installed on the first terminal device 101, the second terminal device 102, or the third terminal device 103, such as shopping applications, web browser applications, search applications, instant messaging tools, email clients, social platform software, financial consumption applications, etc. (for example only).

[0071] The first terminal device 101, the second terminal device 102, and the third terminal device 103 can be various electronic devices with display screens and web browsing support, including but not limited to smartphones, tablet computers, laptop computers, and desktop computers. The first terminal device 101, the second terminal device 102, and the third terminal device 103 can also be referred to as clients. The first terminal device 101, the second terminal device 102, and the third terminal device 103 can be used for functions such as delayed access to the clearing system and file maintenance.

[0072] Server 105, acting as a node in the blockchain, provides storage space and computing power for the entire blockchain. It can be a server providing various services, such as supporting delayed settlement requests sent by users using the first terminal device 101, the second terminal device 102, and the third terminal device 103, and a backend management server for delayed settlement (for example only). The backend management server can perform delayed settlement processing on received merchant delayed settlement requests and provide feedback to the terminal device regarding the processing results (e.g., tables, information, or data obtained or generated based on user requests).

[0073] It should be noted that the delayed settlement method based on blockchain and deep learning provided in the embodiment of the present application can generally be executed by the server 105. Accordingly, the delayed settlement device based on blockchain and deep learning provided in the embodiment of the present application can generally be set in the server 105. The delayed settlement method based on blockchain and deep learning provided in the embodiment of the present application can also be executed by a server or server cluster that is different from the server 105 and can communicate with the first terminal device 101, the second terminal device 102, the third terminal device 103 and / or the server 105. Accordingly, the delayed settlement device based on blockchain and deep learning provided in the embodiment of the present application can also be set in a server or server cluster that is different from the server 105 and can communicate with the first terminal device 101, the second terminal device 102, the third terminal device 103 and / or the server 105.

[0074] It should be understood that Figure 1 The number of terminal devices, networks, servers and databases in the embodiment is merely illustrative. Any number of terminal devices, networks, servers and databases may be provided as required.

[0075] It should be noted that the delayed settlement method and device based on blockchain and deep learning in the embodiments of the present application can be applied to the field of financial technology. The bank's acquiring business has shifted its resources from pursuing the number of acquiring merchants to pursuing the quality of acquiring merchants. For large-scale merchants such as shopping malls, upgrading the acquiring system to support functions such as delayed settlement will help improve the customer stickiness of large-scale merchants and enhance the brand loyalty of large merchants. For merchants who are interested in accessing the delayed settlement function, they are generally merchants with relatively large qualifications and are in a leading position in the industry. Business personnel can invest a lot of marketing resources to strive for them in a targeted manner, which makes a relatively large contribution to the bank's acquiring flow and rebate income; while for merchants who are generally not interested in accessing the delayed settlement function, they are generally merchants with relatively small qualifications and have limited contributions to the bank's acquiring flow and rebate income. The delayed settlement method based on blockchain and deep learning in this application can accurately and efficiently identify merchants with delayed settlement authority and can efficiently execute delayed settlement.

[0076] It should be noted that the delayed settlement method and device based on blockchain and deep learning in the embodiments of the present application can be used for scenarios such as delayed settlement in the financial field, and can also be used for delayed settlement in any field other than the financial field, such as e-commerce, logistics, etc. The application field of the delayed settlement method and device based on blockchain and deep learning in this application is not limited.

[0077] The following will be based on Figure 1 The scene described by Figures 2 to 5 The delayed settlement method based on blockchain and deep learning according to an embodiment of the present application is described in detail.

[0078] Figure 2 The figure schematically shows a flowchart of a delayed settlement method based on blockchain and deep learning according to an embodiment of the present application.

[0079] like Figure 2 As shown, the delayed settlement method based on blockchain and deep learning of this embodiment includes operations S210 to S250, and the delayed settlement method based on blockchain and deep learning can be executed on server 105.

[0080] In operation S210, a transaction proposal sent by a client is received, the transaction proposal is verified by an endorsement node of the blockchain, and signature verification information is generated.

[0081] In operation S220 , signature verification information is sent to the client, and merchant information of multiple merchants sent by the client is received.

[0082] In operation S230 , merchant information of a plurality of merchants is pre-processed and input into a pre-trained neural network model respectively, and the delayed settlement authority of each merchant is output.

[0083] In operation S240, the smart contract is called to store the profile information of the merchant whose delayed settlement authority is allowed to access.

[0084] In operation S250 , in response to the deferred settlement request of the target merchant sent by the client, deferred settlement is performed for the target merchant based on pre-stored profile information.

[0085] According to embodiments of the present application, a transaction proposal may include a smart contract identifier, a smart contract method, parameter information for invoking the smart contract, and client signature information. The smart contract identifier may be the specific smart contract method used for delayed settlement. The smart contract method may be the logic for a specific time limit for delayed settlement. The parameter information for invoking the smart contract may include the required input items for invoking the corresponding smart contract and a flag that uniquely identifies the smart contract. The client signature information may include an identifier that identifies the transaction channel the client is in and a tag for identity verification.

[0086] According to an embodiment of the present application, multiple merchants may be merchants whose signature verification information meets preset conditions. Satisfying the preset conditions may be that the results of the transaction proposals are consistent and are executed with reference to a specified endorsement policy.

[0087] According to an embodiment of the present application, the neural network model includes an input layer, a hidden layer, and an output layer. The input layer is used to input merchant information, and the hidden layer is used to process merchant information. The number of hidden layers and the number of neurons in each hidden layer are related to the number of variables contained in the merchant information. In other words, the structure of the hidden layer of the neural network model is dynamically determined based on the variables in the merchant information.

[0088] According to an embodiment of the present application, merchant information may include, for example, the merchant's monthly turnover, any credit card delinquencies, any financial overdue payments, any fraudulent activity, total monthly tax payments, the merchant's current acquiring institution commission income, daily acquiring institution commission expenditure, percentage of returned transactions, and monthly total amount of returns. Archival information may include clearing account numbers, clearing name information, and delayed clearing periods.

[0089] According to an embodiment of the present application, the merchant's delayed settlement authority can be understood as whether the merchant is qualified to access the delayed settlement system for delayed settlement.

[0090] Through the delayed settlement method of the embodiment of the present application, the number of layers of the hidden layer of the neural network model and the number of neurons in each hidden layer are dynamically determined based on the number of variables contained in the merchant information. The identification of delayed settlement authority that can match different merchant information avoids underfitting and overfitting in the process of identifying delayed settlement authority for merchants, and can accurately identify target merchants, thereby improving the accuracy of deferred settlement. Different merchants use neural network models with corresponding parameters to better utilize computing resources, improve the utilization rate of computing resources and improve the recognition efficiency of the model, thereby improving the efficiency of delayed settlement. The introduction of blockchain technology for delayed settlement can improve the security of data involved in delayed settlement due to the decentralization of blockchain.

[0091] Figure 3 The following schematically illustrates a flow chart of a delayed settlement transaction implemented by an application client interacting with a blockchain according to an embodiment of the present application.

[0092] like Figure 3 As shown, in an embodiment of the present application, a feasible way for the application client to interact with the blockchain to implement a delayed settlement transaction process is:

[0093] The application client (for functions such as access to the delayed clearing system and file maintenance) calls the member service through the SDK to register and obtain the identity certificate for the relevant interfaces of the delayed clearing system.

[0094] The application client initiates a transaction proposal to the blockchain network through the SDK. The transaction proposal sends information such as the smart contract identifier to be called for this transaction (specifically which smart contract method of the delayed settlement system will be used), the contract method (the specific implementation logic of the smart contract method), parameter information (the required input items for calling the corresponding smart contract and the input flag that can uniquely identify the smart contract), and the client signature (an identifier that can identify which transaction channel the client is in and a label for verifying identity) to the endorsing node.

[0095] After receiving the transaction signature, the endorsing node (referring to the Peer node) verifies the signature and confirms whether the submitter has the authority to perform the operation. At the same time, it simulates the execution of the smart contract according to the endorsement policy and returns the result and its respective CA signature to the application client.

[0096] After receiving the information returned by the endorsement node, the application client determines whether the proposal results are consistent and whether to execute according to the specified endorsement policy. Depending on the smart contract being executed, it determines whether it is an onboarding process, a profile entry process, or a delayed settlement accounting process. For the onboarding process, merchant information is entered, including monthly turnover, any credit card delinquencies, any financial overdues, any fraud, monthly tax payments, the merchant's current acquirer commission income, daily acquirer commission expenditure, percentage of returned transactions, and monthly total return amount. Data is preprocessed, and a neural network model is used to perform binary classification on the preprocessed data. Once the merchant meets the access requirements for the delayed settlement system, the client enters profile information and invokes a smart contract to maintain the profile information on the blockchain. Delayed settlement is then performed based on the profile information. The application client packages the data into a signed transaction and sends it to the ordering role (the Order node).

[0097] The sorting role performs consensus sorting on the received transactions, and finally packages a batch of transactions together according to the block generation strategy, generates a new block, and sends it to the submitting node (referring to the Peer node).

[0098] After receiving the block, the commit node verifies each transaction in the block, checks whether the input and output that the transaction depends on are consistent with the current state of the blockchain, and appends the block to the local blockchain and modifies the world state (the latest values ​​of all keys).

[0099] In an embodiment of the present application, verification of transaction proposals by endorsing nodes on a blockchain can be implemented using a modular component abstracted from the distributed ledger. This modular component is used to define the rules for identity verification, authentication, and granting access to the network, and may include the following functions:

[0100] Manage user IDs: A mechanism for managing user IDs that allow access to delayed liquidation identity verification and delayed liquidation processing related functions.

[0101] Verify nodes that want to join the blockchain network: Each node that wants to join the network must provide its valid and legal signature information

[0102] Providing credentials for transactions initiated by clients: Data transmission between nodes (Client, Peer, Orderer) requires verification of the signatures of each node.

[0103] Figure 4 The following schematically shows a logical structure diagram of a modular component according to an embodiment of the present application.

[0104] like Figure 4 As shown, the logical functions of the modular components may include:

[0105] RCA refers to the root CA. The folder contains a list of the root CA's self-signed X.509 certificates, which are used for self-signing and signing intermediate CA certificates.

[0106] ICA refers to Intermediate CA and contains a list of certificates issued by the root CA.

[0107] OU refers to an organizational unit and contains a list of organizational units whose members are considered part of the organization represented by this modular component.

[0108] B refers to Administrators. This folder contains a list of identities that define the roles that have the administrator role for this organization. For standard modular component types, this list should have one or more X.509 certificates.

[0109] ReCA refers to the revocation certificate, which saves the information of the revoked participants. The business user accounts that have withdrawn from the delayed clearing function should be in the revocation certificate list and cannot use related functions.

[0110] SCA refers to the signature certificate, which endorses the signature certificate of the node in the transaction proposal response. This folder is required for local modular components, and the node must have only one X.509 certificate.

[0111] KeyStore refers to private keys. This folder is defined for local modular components of a peer or orderer node and contains the node's signing key. This key cryptographically matches the signing certificate contained in the SCA folder and is used to sign data. This folder is required for local modular components and must contain only one private key.

[0112] TLS Root CAs: Contains a list of self-signed X.509 certificates for root CAs that your organization trusts for TLS communications. There must be at least one TLS root CA X.509 certificate in this folder.

[0113] TLS Intermediate CA: Saves the list of intermediate certificates issued by the TLS root CA.

[0114] Based on the above method, for different transaction parties or merchants, only users with relevant private keys have the authority to query relevant details, which greatly improves the security of user data.

[0115] In an embodiment of the present application, pre-processing of merchant information of multiple merchants may include:

[0116] The merchant information of multiple merchants is standardized and converted into digital data in a preset range.

[0117] For example, you can use the min-max method, often called feature normalization, to ensure that all normalized data falls between [0, 1]. Before applying normalization, you must first calculate the minimum and maximum values ​​for each column in the dataset. Furthermore, for yes / no questions like whether there is fraud, whether there is overdue payment, whether there is a credit card default, and whether the system can pass the delayed clearing system access process, convert "no" to 0 and "yes" to 1.

[0118] Through the delayed settlement method of the embodiment of the present application, the merchant information is subjected to standardized feature processing before being input into the neural network model, and the user information is converted into a digital type, which can solve the pain point problem that non-numeric types cannot be input into the neural network deep learning framework, thereby enabling more accurate identification of target merchants.

[0119] In an embodiment of the present application, after pre-processing merchant information of multiple merchants, the pre-trained neural network model is inputted respectively to output the delayed settlement authority of each merchant, which may include:

[0120] The merchant information is nonlinearly transformed and compressed through the hidden layer of the neural network model to capture the nonlinear relationship between the features in the merchant information and generate the first abstract features of multiple dimensions.

[0121] The first abstract features of multiple dimensions are deeply fused and compressed through the hidden layer of the neural network model to generate the second abstract features.

[0122] A tag of the merchant is determined based on the second abstract feature, where the tag is used to indicate the merchant's delayed settlement authority.

[0123] According to the embodiments of this application, the hidden layer uses multiple neurons to perform nonlinear transformations on the input features. Each neuron uses the ReLU activation function (f(x) = max(0, x)), mapping the linear combination of inputs to nonlinear outputs. The purpose of nonlinear transformation is to capture nonlinear relationships between features (such as the interaction between transaction flow and credit score) and generate higher-dimensional abstract feature representations.

[0124] Multiple neurons compress the original features into multiple new feature dimensions, reducing redundant information while retaining key features. For example, if there are 10 original features, the hidden layer converts the 10-dimensional input into a 5-dimensional output through the weight matrix and ReLU activation.

[0125] The output of the hidden layer serves as the input of the hidden layer, passing the nonlinear features to the next hidden layer for further processing.

[0126] The hidden layer uses multiple neurons to deeply fuse multiple hidden layer features, further extracting high-level abstract features. Each neuron receives the output of all other neurons and generates new features through a weight matrix and ReLU activation. This function captures the complex relationships between features extracted by the hidden layer (such as the combined impact of transaction flow, credit score, and compliance record).

[0127] The multi-dimensional features of the hidden layer are compressed to further focus on the features that are most important for the classification task. For example, if the features output by the hidden layer are 5-dimensional, the hidden layer converts the 5-dimensional input into a 3-dimensional output through the weight matrix and ReLU activation.

[0128] The output features of the hidden layer are highly correlated with whether the merchant meets the entry requirements, providing a key basis for the final classification.

[0129] The hidden layer performs a preliminary nonlinear transformation on the original features and further abstracts the output to form a hierarchical feature representation. The superposition of two hidden layers through the ReLU activation function significantly enhances the model's nonlinear expression capabilities, enabling it to capture more complex feature relationships.

[0130] The output layer of the neural network model uses a single neuron and a Sigmoid activation function to map the output to a probability value in the [0, 1] range. When performing a threshold determination, if the probability value is ≥ 0.5, the merchant is considered to meet the entry requirements; otherwise, it is not.

[0131] Through the delayed settlement method of the embodiment of the present application, the improved neural network performs hierarchical feature extraction on merchant information, so that the neural network model can efficiently process merchant information and accurately determine whether it meets the access conditions for accessing the delayed settlement system, thereby improving the accuracy of delayed settlement.

[0132] In an embodiment of the present application, the number of hidden layers is determined by rounding the logarithmic operation result of the number of variables included in the merchant information, and the number of neurons is determined based on the number of variables included in the merchant information and the number of hidden layers.

[0133] For example, let the number of variables of merchant information be N, then the number of hidden layers L can be the value of log2N+1 rounded up, and the number of neurons can be .

[0134] For example, the input variables involved in merchant information may be 9 variables, so the total number of neurons that can be set can be less than 9. To improve the accuracy of the neural network model in identifying merchant information, the total number of neurons can be set to 8. The number of hidden layers in the binary classification model can be determined by the number of categories. The improved neural network model has two hidden layers, the first hidden layer can include, for example, 5 neurons, and the second hidden layer can include, for example, 3 neurons. It should be understood that this setting is exemplary, and the number of hidden layers and neurons corresponding to different numbers of variables can be different.

[0135] Through the delayed settlement method of the embodiment of the present application, the specific structure of the hidden layer is reasonably set for the characteristic data of the merchant information of the delayed settlement. This structure can better match the delayed settlement and improve the accuracy and efficiency of the delayed settlement.

[0136] In an embodiment of the present application, the data stored on the blockchain can be used as input data for training the neural network model. The above information is entered through the client. After the neural network smart contract is trained, the neural network model will output the predicted result of allowing access or not allowing access, and return the predicted result to the client. In order to predict the accuracy of the neural network model, this device uses a confusion matrix to divide the training set and test set from the blockchain into a ratio of 7:3. The output confusion results are shown in Table 1:

[0137] Table 1

[0138]

[0139] From the results of the confusion matrix, it can be seen that the improved neural network model has a very high accuracy and can be actually applied to whether to allow merchants to enter delayed settlement, with an accuracy of 98.82%.

[0140] Subsequent access / non-access data will also be updated to the blockchain in a timely manner. This data can be used to optimize the performance of the improved neural network model.

[0141] If the blockchain determines that the merchant is qualified to access the delayed settlement system, subsequent processing will continue. Otherwise, the client will be returned to the following message: "This merchant is not recommended to access the delayed settlement system. Please do not access this delayed settlement system." If access is allowed, the client will automatically jump to the front-end file information entry interface. The information required to enter includes: merchant settlement account number, delayed settlement date, merchant number, merchant name, etc. After receiving the file information entered by the front-end system, the blockchain system calls the smart contract to store the relevant information on the blockchain, using the merchant number as the unique key and the remaining values ​​as values, to ensure that the relevant information cannot be tampered with.

[0142] In an embodiment of the present application, performing delayed settlement for a target merchant based on pre-stored profile information includes:

[0143] Merchant consumption funds within the delayed settlement date will be temporarily stored in an internal account.

[0144] In response to a customer's return request, the smart contract is called to deduct the relevant funds from the internal account and return them to the customer, and the fund information stored in the blockchain is simultaneously updated.

[0145] For example, when a merchant connected to the delayed settlement system makes a purchase, the consumption amount within the settlement date range will be retained in the newly added delayed accounting and write-off internal account according to the delayed settlement date set in the file; when processing return requests, the funds in the newly added internal account can be used for netting. If the return amount is greater than the amount in the delayed accounting and write-off internal account, the return will fail and the error "Funds to be recorded are less than the return amount" will be reported.

[0146] Through the delayed settlement method of the embodiment of the present application, by setting up an internal account in the blockchain, the smart contract can deduct the relevant funds from the internal account and return them to the customer, thereby improving the efficiency of account settlement.

[0147] In an embodiment of the present application, performing delayed settlement for the target merchant based on pre-stored profile information may further include:

[0148] For merchants who have exceeded the liquidation date range and have any surplus, the smart contract is called to deduct the relevant funds from the internal account for the merchant to credit.

[0149] When the funds in the internal account exceed the preset threshold, the excess funds will be stored in the designated account and the fund information stored in the blockchain will be updated synchronously.

[0150] For example, when the excess funds exceed 30% of the funds to be liquidated, the cross-chain circuit breaker protocol is activated, the excess funds are stored in the supervisory account of the execution, and an alarm message is issued. The fund information stored in the blockchain is updated synchronously, and the blockchain's smart contract is called to generate a final report on the flow of funds.

[0151] The delayed settlement method of this embodiment of the application establishes an internal account in the blockchain, allowing the smart contract to deduct relevant funds from the internal account and return them to the merchant, improving account settlement efficiency. Furthermore, the addition of a cross-chain circuit breaker protocol enables risk prevention and control, improving account security.

[0152] In an embodiment of the present application, the method further includes:

[0153] In response to deferred settlement requests from different target merchants sent by the client, the deferred settlement requests from different target merchants are allocated to different processing channels in the blockchain, and deferred settlement is performed for different target merchants based on pre-stored profile information.

[0154] For example, different channels can be established based on the function of distributed ledgers, allowing for delayed settlement and creating a separate ledger for transactions. Only participants in the same channel will have the ledger for that channel, while other participants not in the channel cannot see this ledger. This channel isolation technology provides higher security. In an embodiment of the present application, the method further includes:

[0155] The updated blockchain data is input into the pre-trained neural network model, and the parameters of the pre-trained neural network model are updated.

[0156] According to embodiments of the present application, real-time delayed settlement data can be transmitted back to the blockchain as training data. This data is then fed into a neural network model to optimize its parameters, achieving a closed loop between real-time feedback from the blockchain and the dynamic structure neural network. This improves the model's prediction accuracy, and thus the accuracy of the delayed settlement method.

[0157] In an embodiment of the present application, the delayed settlement authority output by the neural network is compared with the stored merchant historical delayed settlement information to obtain the deviation predicted by the model. When the deviation is greater than a preset threshold, multiple nodes in the blockchain are triggered to perform voting arbitration to ensure the accuracy of delayed settlement.

[0158] Figure 5 The figure schematically illustrates a delayed settlement method based on blockchain and deep learning according to another embodiment of the present application.

[0159] like Figure 5 As shown, the blockchain and deep learning-based delayed settlement method of this embodiment is applied to the client, and the method includes operations S510 to S540.

[0160] In operation S510, signature verification information sent by an endorsing node of the blockchain is received, where the signature verification information is generated by the endorsing node based on the transaction proposal.

[0161] The transaction proposal includes the smart contract identifier, contract method, parameter information for calling the smart contract, and the client's signature information.

[0162] In operation S520, the merchant information of the merchant whose signature verification information meets the preset conditions is entered and sent to the blockchain, so that the blockchain pre-processes the merchant information of multiple merchants and inputs them into the pre-trained neural network model respectively, and outputs the delayed settlement authority of each merchant.

[0163] The number of hidden layers and the number of neurons in each hidden layer used to process merchant information in the neural network model are related to the number of variables contained in the merchant information.

[0164] In operation S530, the profile information of the merchant whose delayed settlement authority is allowed to be accessed is sent to the blockchain for storage.

[0165] In operation S540 , in response to the target merchant inputting a deferred settlement request, the deferred settlement request is sent to the blockchain, so that the blockchain performs delayed settlement for the target merchant based on pre-stored profile information.

[0166] It should be noted that for the details not yet completed in the delayed settlement method of this embodiment and the technical effects brought about, please refer to the aforementioned embodiment section and will not be repeated here.

[0167] In an embodiment of the present application, the delayed settlement method may further include: constructing a visual interface. In response to the merchant opening the visual interface through the client browser, the merchant's merchant information and profile information are entered through the visual interface, and the delayed settlement result is displayed to the merchant.

[0168] According to the embodiments of this application, applications implementing delayed settlement can be managed in a hierarchical manner, with the addition of a control layer and a view layer. The view layer provides a visual interface and interaction for users; the control layer receives delayed settlement requests, accesses the business layer, and then invokes the user chain code to perform data preprocessing, access identity access to the delayed settlement system, enter profile information, access the delayed settlement system, and, in the event of a return, implement real-time return through transitional internal account netting and other smart contracts.

[0169] Through the delayed settlement method of the embodiment of the present application, by layering the application and providing a visual interface on the client, merchants can use the browser to enter identity verification materials of merchants intending to delay settlement and delayed settlement business file information. The operation is convenient, and the front-end and back-end are layered, which improves the scalability and maintainability of the application.

[0170] It should be noted that the delayed settlement method of the embodiment of the present application can serve high-potential merchants with large service providers. It uses a neural network model to determine whether the merchant has relevant intentions, which helps to preliminarily screen high-potential merchants connected to the system and improve the stickiness of high-potential merchants by supporting special payment processing. The delayed settlement method of the embodiment of the present application provides convenience for high-potential merchants. It uses the amount that will be directly credited to the merchant the next day to determine the delayed date set in the file, and stores the funds on the delayed date into the bank's internal account. When the merchant has a return demand later, the funds can be returned to the customer directly from the internal account, which improves the efficiency of return processing. If the funds in the internal account are lower than the return funds, the return fails and an error message is given: "The funds to be credited are less than the return amount." The newly added internal account for the transitional funds for delayed POS netting returns is used to delay the advance payment on the day of clearing, and then automatically write off the relevant funds in batches. At the same time, it should be noted that since large group merchants have sufficient funds, delayed accounting will not have much impact on them. However, responding quickly to customer returns will help improve customer satisfaction with large group merchants, thereby increasing the revenue of large group merchants. It will also help to deeply bind large group merchants with system providers, help stabilize banks' collection flows and increase banks' commission income.

[0171] Based on the above-mentioned delayed settlement method based on blockchain and deep learning, this application also provides a delayed settlement device based on blockchain and deep learning. Figure 8 The device is described in detail.

[0172] Figure 6 The structural block diagram of a delayed settlement device based on blockchain and deep learning according to an embodiment of the present application is schematically shown.

[0173] like Figure 6 As shown, the delayed settlement device 600 based on blockchain and deep learning of this embodiment includes a first receiving module 610, a sending and receiving module 620, a preprocessing module 630, a calling module 640 and a first response module 650.

[0174] The first receiving module 610 is configured to receive a transaction proposal sent by a client, verify the transaction proposal through an endorsing node on the blockchain, and generate signature verification information. The transaction proposal includes a smart contract identifier, a contract method, parameter information for invoking the smart contract, and the client's signature information. In one embodiment, the first receiving module 610 can be configured to perform operation S210 described above and will not be further described here.

[0175] The sending and receiving module 620 is configured to send signature verification information to the client and receive merchant information of multiple merchants sent by the client, wherein the multiple merchants are merchants whose signature verification information meets preset conditions. In one embodiment, the sending and receiving module 620 can be used to perform the operation S220 described above, which will not be repeated here.

[0176] Preprocessing module 630 is configured to preprocess merchant information for multiple merchants and then input it into a pretrained neural network model to output each merchant's delayed settlement authority. The number of hidden layers and the number of neurons in each hidden layer used to process merchant information in the neural network model are related to the number of variables contained in the merchant information. In one embodiment, preprocessing module 630 can be used to perform operation S230 described above and will not be further described here.

[0177] The calling module 640 is used to call the smart contract to store the profile information of the merchant whose delayed liquidation authority is allowed to access. In one embodiment, the calling module 640 can be used to perform the operation S240 described above, which will not be repeated here.

[0178] The first response module 650 is configured to respond to the target merchant's deferred settlement request sent by the client and perform deferred settlement for the target merchant based on pre-stored profile information. In one embodiment, the first response module 650 may be configured to perform the aforementioned operation S250, which will not be described in detail herein.

[0179] According to an embodiment of the present application, the pre-processing module 630 pre-processes the merchant information of multiple merchants, which may include: performing feature standardization on the merchant information of the multiple merchants, and converting the merchant information into digital data of a preset interval.

[0180] According to an embodiment of the present application, after pre-processing the merchant information of multiple merchants, the pre-trained neural network model is inputted respectively to output the delayed settlement authority of each merchant. This may include: performing nonlinear transformation and compression on the merchant information through the hidden layer of the neural network model to capture the nonlinear relationship between the features in the merchant information and generate a first abstract feature of multiple dimensions. Deeply fusing and compressing the first abstract features of multiple dimensions through the hidden layer of the neural network model to generate a second abstract feature. Based on the second abstract feature, the merchant's label is determined, and the label is used to indicate the merchant's delayed settlement authority.

[0181] According to an embodiment of the present application, the hidden layer includes 5 neurons, and the hidden layer includes 3 neurons.

[0182] According to an embodiment of the present application, the first response module 650 performs delayed settlement for the target merchant based on pre-stored profile information, including temporarily depositing merchant consumption funds within the delayed settlement date into an internal account. In response to a customer-initiated return request, the smart contract is invoked to deduct the relevant funds from the internal account and return them to the customer, while simultaneously updating the funds information stored on the blockchain.

[0183] According to an embodiment of the present application, the first response module 650 performs delayed liquidation for the target merchant based on pre-stored archival information, and also includes: for merchants that exceed the liquidation date range and have any surplus, calling the smart contract to deduct relevant funds from the internal account for the merchant to deposit.

[0184] According to an embodiment of the present application, the device also includes: a channel establishment module 660, which is used to respond to deferred settlement requests from different target merchants sent by the client, allocate deferred settlement requests from different target merchants to different processing channels in the blockchain, and perform delayed settlement for different target merchants based on pre-stored archival information.

[0185] Figure 7 The structural block diagram of a delayed settlement device based on blockchain and deep learning according to another embodiment of the present application is schematically shown.

[0186] like Figure 7 As shown, the delayed settlement device 700 based on blockchain and deep learning of this embodiment includes a second receiving module 710, an input module 720, a first sending module 730 and a second sending module 740.

[0187] The second receiving module 710 is configured to receive signature verification information sent by an endorsing node of the blockchain. The signature verification information is generated by the endorsing node based on a transaction proposal. The transaction proposal includes a smart contract identifier, a contract method, parameter information for invoking the smart contract, and the client's signature information. In one embodiment, the second receiving module 710 can be configured to perform operation S510 described above and will not be further described here.

[0188] Input module 720 is used to input merchant information of merchants whose signature verification information meets preset conditions and send it to the blockchain. The blockchain then pre-processes the merchant information of multiple merchants and inputs it into a pre-trained neural network model to output the delayed settlement permissions of each merchant. The number of hidden layers and the number of neurons in each hidden layer used to process the merchant information in the neural network model are related to the number of variables included in the merchant information. In one embodiment, input module 720 can be used to perform operation S520 described above and will not be further described here.

[0189] The first sending module 730 is configured to send the profile information of the merchant whose delayed settlement authority is allowed to be accessed to the blockchain for storage. In one embodiment, the first sending module 730 can be used to perform the operation S530 described above, which will not be repeated here.

[0190] Second sending module 740 is configured to respond to the target merchant's input of a deferred settlement request and send the deferred settlement request to the blockchain, so that the blockchain performs deferred settlement for the target merchant based on pre-stored profile information. In one embodiment, second sending module 740 can be configured to perform operation S540 described above, which will not be further described here.

[0191] It should be noted that the specific implementation details and technical effects of the device embodiment can be found in the method embodiment part, which will not be repeated here.

[0192] According to embodiments of the present application, any multiple modules among the first receiving module 610, the sending and receiving module 620, the pre-processing module 630, the calling module 640, and the first response module 650 may be combined into a single module, or any one of these modules may be split into multiple modules. Alternatively, at least part of the functionality of one or more of these modules may be combined with at least part of the functionality of other modules and implemented in a single module. According to embodiments of the present application, at least one of the first receiving module 610, the sending and receiving module 620, the pre-processing module 630, the calling module 640, and the first response module 650 may be at least partially implemented as a hardware circuit, such as a field programmable gate array (FPGA), a programmable logic array (PLA), a system on a chip, a system on a substrate, a system on a package, an application-specific integrated circuit (ASIC), or may be implemented in hardware or firmware through any other reasonable means of circuit integration or packaging, or may be implemented in any one of the three implementation methods of software, hardware, and firmware, or any appropriate combination of any of these. Alternatively, at least one of the first receiving module 610, the sending and receiving module 620, the preprocessing module 630, the calling module 640 and the first responding module 650 can be at least partially implemented as a computer program module, which can perform corresponding functions when executed.

[0193] Figure 8 A block diagram of an electronic device suitable for implementing a delayed settlement method based on blockchain and deep learning according to an embodiment of the present application is schematically shown.

[0194] like Figure 8As shown, an electronic device 800 according to an embodiment of the present application includes a processor 801, which can perform various appropriate actions and processes based on a program stored in a read-only memory (ROM) 802 or a program loaded from a storage unit 808 into a random access memory (RAM) 803. The processor 801 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and / or a related chipset and / or a dedicated microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 801 may also include onboard memory for caching purposes. The processor 801 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of the present application.

[0195] Various programs and data required for the operation of the electronic device 800 are stored in the RAM 803. The processor 801, ROM 802, and RAM 803 are connected to each other via a bus 804. The processor 801 performs various operations of the method flow according to the embodiment of the present application by executing the programs in the ROM 802 and / or RAM 803. It should be noted that the programs may also be stored in one or more memories other than the ROM 802 and the RAM 803. The processor 801 may also perform various operations of the method flow according to the embodiment of the present application by executing the programs stored in the one or more memories.

[0196] According to an embodiment of the present application, electronic device 800 may further include an input / output (I / O) interface 805, which is also connected to bus 804. Electronic device 800 may also include one or more of the following components connected to I / O interface 805: an input section 806 including a keyboard, mouse, etc.; an output section 807 including devices such as a cathode ray tube (CRT), liquid crystal display (LCD), and speakers; a storage section 808 including a hard disk; and a communication section 809 including a network interface card such as a LAN card or modem. Communication section 809 performs communication processing via a network such as the Internet. A drive 810 is also connected to I / O interface 805 as needed. Removable media 811, such as a magnetic disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed in drive 810 as needed, so that computer programs read from the removable media can be installed into storage section 808 as needed.

[0197] This application also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiments, or may exist independently and not be incorporated into the device / apparatus / system. The computer-readable storage medium carries one or more programs, and when the one or more programs are executed, the method according to the embodiments of this application is implemented.

[0198] According to an embodiment of the present application, a computer-readable storage medium may be a non-volatile computer-readable storage medium, and may include, for example, but not limited to: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present application, a computer-readable storage medium may be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. For example, according to an embodiment of the present application, a computer-readable storage medium may include the ROM 802 and / or RAM 803 described above and / or one or more memories other than ROM 802 and RAM 803.

[0199] Embodiments of the present application also include a computer program product comprising a computer program containing program code for executing the method shown in the flowchart. When the computer program product is executed in a computer system, the program code is used to cause the computer system to implement the method provided in the embodiments of the present application.

[0200] The computer program executes the above functions defined in the system / device of the embodiment of the present application when the processor 801 executes the computer program. According to the embodiment of the present application, the system, device, module, unit, etc. described above can be implemented by a computer program module.

[0201] In one embodiment, the computer program may be stored on a tangible storage medium such as an optical storage device or a magnetic storage device. In another embodiment, the computer program may be transmitted and distributed in the form of a signal on a network medium, downloaded and installed via the communication portion 809, and / or installed from a removable medium 811. The program code contained in the computer program may be transmitted using any appropriate network medium, including but not limited to wireless, wired, or any suitable combination thereof.

[0202] In such an embodiment, the computer program can be downloaded and installed from a network via the communication section 809, and / or installed from a removable medium 811. When the computer program is executed by the processor 801, the above-described functions defined in the system of the embodiment of the present application are performed. According to the embodiment of the present application, the systems, devices, means, modules, units, etc. described above can be implemented by computer program modules.

[0203] According to an embodiment of the present application, the program code for executing the computer program provided by the embodiment of the present application can be written in any combination of one or more programming languages. Specifically, these computer programs can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. Programming languages ​​include, but are not limited to, languages ​​such as Java, C++, Python, "C" or similar programming languages. The program code can be executed entirely on the user computing device, partially on the user device, partially on a remote computing device, or entirely on a remote computing device or server. In the case of a remote computing device, the remote computing device can be connected to the user computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computing device (for example, using an Internet service provider to connect via the Internet).

[0204] The flowcharts and block diagrams in the accompanying drawings illustrate the possible implementation architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present application. In this regard, each box in the flowchart or block diagram can represent a module, program segment, or a part of code, and the above-mentioned module, program segment, or a part of code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in an order different from that marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flowchart, and the combination of the boxes in the block diagram or flowchart, can be implemented with a dedicated hardware-based system that performs the specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0205] Those skilled in the art will appreciate that the features described in the various embodiments of this application may be combined and / or coupled in various ways, even if such combinations or couplings are not explicitly described in this application. In particular, the features described in the various embodiments of this application may be combined and / or coupled in various ways without departing from the spirit and teachings of this application. All such combinations and / or couplings fall within the scope of this application.

Claims

1. A delayed settlement method based on blockchain and deep learning, characterized in that: Applied to blockchain, the method includes: Receive transaction proposals sent by clients, verify the transaction proposals through the blockchain’s endorsement nodes, and generate signature verification information. The transaction proposals include the smart contract identifier, smart contract method, parameter information for calling the smart contract, and the client’s signature information. Sending the signature verification information to the client, and receiving merchant information of multiple merchants sent by the client, wherein the multiple merchants are merchants whose signature verification information meets preset conditions; After pre-processing merchant information of multiple merchants, the pre-trained neural network model is inputted respectively, and the delayed settlement authority of each merchant is outputted. The number of hidden layers and the number of neurons in each hidden layer used to process the merchant information in the neural network model are dynamically determined based on the number of variables contained in the merchant information. Calling the smart contract to store the profile information of merchants whose delayed settlement authority is allowed to access; and In response to a deferred settlement request from a target merchant sent by a client, deferred settlement is performed for the target merchant based on pre-stored profile information.

2. The method according to claim 1, characterized in that The pre-processing of merchant information of multiple merchants includes: The merchant information of multiple merchants is subjected to feature standardization, and the merchant information is converted into digital data of a preset range.

3. The method according to claim 1, characterized in that The merchant information of multiple merchants is pre-processed and then input into the pre-trained neural network model to output the delayed settlement authority of each merchant, including: Performing nonlinear transformation and compression on the merchant information through the hidden layer of the neural network model to capture the nonlinear relationship between features in the merchant information and generate first abstract features of multiple dimensions; Deeply fusing and compressing the first abstract features of multiple dimensions through the hidden layer of the neural network model to generate a second abstract feature; A tag of the merchant is determined based on the second abstract feature, where the tag is used to indicate the delayed settlement authority of the merchant.

4. The method according to claim 3, characterized in that The number of layers of the hidden layer is determined by rounding the result of a logarithmic operation on the number of variables included in the merchant information, and the number of neurons is determined based on the number of variables included in the merchant information and the number of layers of the hidden layer.

5. The method according to claim 1, wherein The performing delayed settlement for the target merchant based on pre-stored profile information includes: Temporarily store merchant consumption funds within the delayed settlement date in an internal account; In response to a return request initiated by a customer, the smart contract is called to deduct the relevant funds from the internal account and return them to the customer, and the fund information stored in the blockchain is simultaneously updated.

6. The method according to claim 5, characterized in that The performing of delayed settlement for the target merchant based on pre-stored profile information further includes: For merchants that have exceeded the liquidation date range and have any surplus, the smart contract is called to deduct the relevant funds from the internal account for the merchant to credit; When the funds in the internal account exceed a preset threshold, the excess funds are stored in a designated account, and the fund information stored in the blockchain is simultaneously updated.

7. The method according to claim 1, characterized in that The method further comprises: In response to deferred settlement requests from different target merchants sent by the client, the deferred settlement requests from different target merchants are allocated to different processing channels in the blockchain, and deferred settlement is performed for different target merchants based on pre-stored profile information.

8. The method according to claim 1, characterized in that The method further comprises: The updated data of the blockchain is input into the pre-trained neural network model, and the parameters of the pre-trained neural network model are updated.

9. A delayed settlement method based on blockchain and deep learning, characterized in that: Applied to a client, the method includes: Receive signature verification information sent by an endorsing node of the blockchain, where the signature verification information is generated by the endorsing node based on a transaction proposal, where the transaction proposal includes a smart contract identifier, a smart contract method, parameter information for invoking the smart contract, and signature information of the client; The merchant information of merchants whose signature verification information meets preset conditions is input and sent to the blockchain, so that the blockchain pre-processes the merchant information of multiple merchants and inputs them into a pre-trained neural network model respectively, outputting the delayed settlement authority of each merchant. The number of hidden layers and the number of neurons in each hidden layer used to process the merchant information in the neural network model are dynamically determined based on the number of variables contained in the merchant information. Sending the entered profile information of merchants whose delayed settlement authority is allowed to access to the blockchain for storage; In response to a target merchant inputting a deferred liquidation request, the deferred liquidation request is sent to the blockchain, so that the blockchain performs deferred liquidation for the target merchant based on pre-stored profile information.

10. The method according to claim 9, characterized in that The method further comprises: Build a visual interface; In response to the merchant opening the visual interface through the browser of the client, the merchant information and file information of the merchant are input through the visual interface, and the result of the delayed settlement is displayed to the merchant.

11. A delayed settlement device for acquiring based on blockchain and deep learning, characterized in that: Applied to blockchain, the device includes: The first receiving module is configured to receive a transaction proposal sent by a client, verify the transaction proposal through an endorsement node of the blockchain, and generate signature verification information. The transaction proposal includes a smart contract identifier, a smart contract method, parameter information for calling the smart contract, and the client's signature information; a sending and receiving module, configured to send the signature verification information to the client, and receive merchant information of multiple merchants sent by the client, wherein the multiple merchants are merchants whose signature verification information meets preset conditions; a preprocessing module, configured to preprocess merchant information of multiple merchants and input the pre-trained neural network model into the pre-trained model to output the delayed settlement authority of each merchant, wherein the number of hidden layers and the number of neurons in each hidden layer used to process the merchant information in the neural network model are dynamically determined based on the number of variables contained in the merchant information; A calling module, configured to call the smart contract to store the profile information of merchants whose delayed settlement authority is allowed to access; and The first response module is configured to respond to a target merchant's deferred settlement request sent by the client, and perform deferred settlement for the target merchant based on pre-stored profile information.

12. A delayed settlement device for acquiring based on blockchain and deep learning, characterized in that: Applied to a client, the device includes: A second receiving module is configured to receive signature verification information sent by an endorsing node of the blockchain, where the signature verification information is generated by the endorsing node based on a transaction proposal, where the transaction proposal includes a smart contract identifier, a smart contract method, parameter information for invoking the smart contract, and signature information of the client; An input module, configured to input merchant information of merchants whose signature verification information satisfies preset conditions and send the information to the blockchain, so that the blockchain pre-processes the merchant information of multiple merchants and inputs each of the information into a pre-trained neural network model to output the delayed settlement authority of each merchant. The number of hidden layers and the number of neurons in each hidden layer used to process the merchant information in the neural network model are dynamically determined based on the number of variables contained in the merchant information. A first sending module is used to send the entered profile information of merchants whose delayed settlement authority is allowed to access to the blockchain for storage; The second sending module is configured to respond to a deferred settlement request input by a target merchant and send the deferred settlement request to the blockchain, so that the blockchain performs deferred settlement for the target merchant based on pre-stored profile information.

13. An electronic device comprising: one or more processors; a memory for storing one or more computer programs, It is characterized in that the one or more processors execute the one or more computer programs to implement the steps of the method according to any one of claims 1 to 10.

14. A computer-readable storage medium having a computer program or instruction stored thereon, characterized in that: When the computer program or instruction is executed by a processor, the steps of the method according to any one of claims 1 to 10 are implemented.

15. A computer program product comprising a computer program or instructions, characterized in that When the computer program or instruction is executed by a processor, the steps of the method according to any one of claims 1 to 10 are implemented.

Images