A unified payment channel management method

Through blockchain technology and smart contract combinations with machine learning algorithms, the problem of insufficient real-time risk management and anti-fraud capabilities in unified management of payment channels is solved, and real-time transaction monitoring and security improvement are achieved.

CN118552198BActive Publication Date: 2024-11-22无锡锡商银行股份有限公司
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411004838.8
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-07-25
Publication Date
2024-11-22
Estimated Expiration
2044-07-25

AI Technical Summary

Technical Problem

The existing unified management of payment channels lacks real-time risk management and anti-fraud capabilities, making it difficult to accurately identify and intercept fraudulent transactions, especially the weak ability to handle suspicious transactions on non-blacklists.

Method used

A distributed ledger is built based on blockchain technology, transaction verification is carried out through decentralized identity verification and smart contracts, transaction data is analyzed in combination with machine learning algorithms, anti-fraud rules and thresholds are set, and abnormal transactions are automatically identified and marked.

Benefits of technology

Real-time risk management and fraud detection of payment channels are realized, reducing manual operation errors and management costs, and improving transaction security and transparency.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118552198B_ABST
    Figure CN118552198B_ABST
Patent Text Reader

Abstract

The present invention discloses a unified management method for payment channels, which specifically relates to the field of payment management technology. The specific steps include building a distributed ledger based on blockchain technology, sharing real-time data for all nodes in the distributed ledger, confirming the identities of transaction parties through a decentralized identity authentication system, verifying transactions using public key infrastructure and digital signature technology, setting up smart contracts, automatically executing transactions according to set anti-fraud rules, and combining machine learning algorithms to analyze transaction data to identify and classify abnormal transactions, and automatically marking abnormal transactions. This application optimizes risk management in unified management of payment channels by using blockchain technology, defines constraints on anti-fraud smart contracts of blockchain based on machine learning algorithms, and combines transparent and automated execution of smart contracts to avoid errors and omissions in manual operations, thereby further reducing the risk management costs of unified payment.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of payment management technology, and more specifically, to a payment channel unified management method. Background Art

[0002] When competing with Internet finance third-party payments, banks have not yet completed the construction of a modern payment system. With the expansion of new payment demands such as payment scenarios, mobile payments, and customer analysis, payment platforms have always been attached to the core systems of banks. It is difficult to improve the defects of the payment platform architecture simply by adding support for third-party payments, especially the key real-time risk management and anti-fraud capabilities in the unified management of payment channels. The existing processing method is to filter against the blacklist. For suspicious transactions not on the blacklist, banks usually use multi-factor authentication and post-notification methods. The real-time interception capability during the transaction is weak, and it is difficult to accurately identify fraudulent transactions.

[0003] In order to solve the above defects, a technical solution is now proposed. Summary of the invention

[0004] The purpose of the present invention is to provide a unified payment channel management method to solve the problems in the above-mentioned background technology.

[0005] To achieve the above object, the present invention provides the following technical solution: a payment channel unified management method, the specific steps include:

[0006] Build a distributed ledger based on blockchain technology and share real-time data among all nodes in the distributed ledger;

[0007] The identities of the transaction parties are confirmed through a decentralized identity verification system, and the transactions are verified using public key infrastructure and digital signature technology;

[0008] Set up smart contracts to automatically execute transactions according to the set anti-fraud rules;

[0009] Combined with machine learning algorithms, by analyzing transaction data, abnormal transactions are identified and classified, and abnormal transactions are automatically marked.

[0010] Preferably, the method for setting up a smart contract is:

[0011] When smart contracts are running on the blockchain, they can collect transaction data, including transaction amount, transaction frequency, transaction time, transaction party identity information, and transaction party geographic location. Rules are defined for smart contracts, and anti-fraud operations are performed using smart contracts. The rules are based on the following factors:

[0012] Detection is performed by setting thresholds, including amount thresholds and frequency thresholds;

[0013] Amount threshold: set the upper limit of a single transaction or the cumulative transaction amount within a period of time;

[0014] Frequency threshold: Set the maximum number of transactions allowed within a certain period of time;

[0015] Identification through transaction patterns, including geographic location and time;

[0016] Geographic location anomalies: detect whether the location where the transaction occurred is inconsistent with the user's usual location;

[0017] Time anomaly: detect whether the transaction occurs during the user's abnormal activity time period;

[0018] Analyze through transaction behavior, including historical comparison and device verification;

[0019] Historical data comparison: Compare with the user's historical transaction behavior to detect abnormal patterns;

[0020] Device information: Check if the transaction is from a commonly used device;

[0021] Filtering through blacklists and whitelists;

[0022] Blacklist: Block transactions with known fraudsters or suspicious accounts;

[0023] Whitelist: Allow transactions with trusted accounts.

[0024] Preferably, a method for verifying transactions based on smart contracts is:

[0025] After the transaction is initiated, the transaction data is submitted to the blockchain and processed by the smart contract. The smart contract verifies the transaction according to the set anti-fraud rules and calculates the transaction fraud index by establishing a logistic regression model. The calculation expression is: In the formula, Th is the threshold rule coefficient, Pr is the pattern recognition coefficient, Ba is the behavior analysis coefficient, and Li is the blacklist filtering coefficient;

[0026] When the transaction party is included in the blacklist, the blacklist filtering coefficient is 0, and when the transaction party is not included in the blacklist, the blacklist filtering coefficient is 1.

[0027] Preferably, the calculation method of the threshold rule coefficient and the pattern recognition coefficient is:

[0028] The calculation expression of the threshold rule coefficient is Th = α*Am+β*Fe, where Am is the transaction amount, Fe is the transaction frequency, α and β are the weight coefficients of the transaction amount and transaction frequency, and both α and β are greater than 0;

[0029] The calculation expression of the pattern recognition coefficient is Pr = γ*Ge + δ*Ti, where Ge is the geographical factor value, Ti is the time factor value, γ and δ are the weight coefficients of the geographical factor value and the time factor value respectively, and both γ and δ are greater than 0;

[0030] When the transaction location is inconsistent with the usual location of the transaction party, the geographical factor value is 0; when the transaction location is consistent with the usual location of the transaction party, the geographical factor value is 1;

[0031] When the transaction occurs outside the user's normal activity time period, the time factor value is 0; when the transaction occurs within the user's normal activity time period, the time factor value is 1.

[0032] Preferably, the logic for obtaining the behavior analysis coefficient is:

[0033] The calculation expression of the behavior analysis coefficient is Ba=E*(Ab+1), where E is the equipment factor value. When the equipment used by the transaction party is not a common equipment, E takes a value of 1. When the equipment used by the transaction party is a common equipment, E takes a value of 0. Ab is the abnormality coefficient. The abnormality coefficient is a parameter for judging fraudulent transactions based on the machine learning model by analyzing and processing the historical transaction data of the transaction party using a machine learning algorithm. When the machine learning model identifies the transaction behavior as a fraud risk transaction, Ab takes a value of 1. When the machine learning model identifies the transaction behavior as a stable and safe transaction, Ab takes a value of 0.

[0034] Preferably, the logic of calculating the abnormal coefficient using the machine learning method is:

[0035] Collect historical transaction data of the transaction parties and classify the historical transaction data, including tendency data, account data, relationship data, and external data. The tendency data includes the user's online frequency, online path, IP address change frequency, and browsing behavior characteristics. The account data includes the user's account type, historical successful transactions, and historical failed transactions. The relationship data includes the transaction connection between the user's accounts and the behavior information of binding social accounts. The external data includes third-party credit scores, regional risk levels, and market volatility information.

[0036] Online access path includes the user's operating system type and online channel; browsing behavior characteristics include browsing interface, dwell time, and click behavior; user account types are divided into personal accounts and collective accounts;

[0037] Establish machine learning models based on tendency data, account data, relationship data, and external data, clean the data, including processing missing values ​​and outliers, standardizing data of different scales, converting categorical data into numerical form, extracting features based on different data, using feature selection algorithms to select features, reducing dimensions to improve model performance, training data using supervised learning methods, optimizing model hyperparameters through grid search or random search, optimizing feature extraction and selection, and using trained models to judge transaction events.

[0038] Preferably, the method for identifying and classifying abnormal transactions and automatically marking abnormal transactions in combination with a machine learning algorithm is:

[0039] Setting a transaction fraud threshold and comparing the calculated transaction fraud index with the transaction fraud threshold;

[0040] When the abnormal coefficient of a transaction event is 1, and the transaction fraud index is greater than or equal to the transaction fraud threshold, the transaction event is marked as a crisis event, the transaction activity is frozen and reported to the management personnel;

[0041] When the abnormal coefficient of a transaction event is 1 and the transaction fraud index is less than the transaction fraud threshold, the transaction event is marked as a risk event, the transaction activity is monitored separately and the management personnel are informed;

[0042] When the abnormal coefficient of a transaction event is 0 and the transaction fraud index is greater than or equal to the transaction fraud threshold, the transaction event is marked as a critical event, the transaction activity is marked, and the management personnel are reminded to verify it;

[0043] When the abnormal coefficient of a transaction event is 0 and the transaction fraud index is less than the transaction fraud threshold, the transaction event is marked as a stable event and no marking operation is performed.

[0044] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0045] This application optimizes the risk management in the unified management of payment channels by using blockchain technology, integrates different payment channels using the distributed architecture of blockchain, records all transactions of payment channels in the same blockchain ledger, and uses the security and confidentiality of blockchain to provide personalized security services for transaction users. It defines constraints on the blockchain's anti-fraud smart contracts based on machine learning algorithms, combines the transparent and automated execution of smart contracts, avoids errors and omissions in manual operations, and further reduces the risk management costs of unified payments. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0047] Figure 1 The figure is a flow chart of the method of the present invention. DETAILED DESCRIPTION

[0048] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0049] Example 1

[0050] See also Figure 1 As shown, the present invention is a unified payment channel management method, and the specific steps include:

[0051] Build a distributed ledger based on blockchain technology and share real-time data among all nodes in the distributed ledger;

[0052] The identities of the transaction parties are confirmed through a decentralized identity verification system, and the transactions are verified using public key infrastructure and digital signature technology;

[0053] Set up smart contracts to automatically execute transactions according to the set anti-fraud rules;

[0054] Combined with machine learning algorithms, by analyzing transaction data, abnormal transactions are identified and classified, and abnormal transactions are automatically marked.

[0055] Blockchain technology is a decentralized, transparent and secure technology with many characteristics that make it have great potential in the unified management of payment channels. Blockchain is a distributed ledger technology with the following key characteristics:

[0056] Decentralization: No centralized third-party organization is required to verify and record transactions.

[0057] Transparency: All participating nodes can access a copy of the ledger, and transaction records are open and transparent.

[0058] Immutability: Once a transaction is recorded on the blockchain, it cannot be tampered with.

[0059] Security: Cryptography is used to ensure the security and privacy of transactions.

[0060] The use of blockchain technology for unified management of payment channels has many advantages over the existing centralized payment channel unified management:

[0061] Multi-channel integration: Blockchain can serve as a unified underlying platform to integrate various payment channels. Transactions of all payment channels are recorded in the same blockchain ledger to achieve data unification and transparency. Different payment channels can achieve seamless interoperability and settlement through smart contracts.

[0062] Enhanced security: The distributed and encrypted nature of blockchain can enhance the security of payment channels, verify transactions through a consensus mechanism, prevent tampering and fraud, and all transaction data is protected by encryption technology to ensure privacy and security.

[0063] Real-time settlement: Blockchain technology can support real-time payment settlement. Through the peer-to-peer transaction characteristics of blockchain, it can realize real-time payment processing and reduce the delay of transaction confirmation by optimizing the consensus algorithm and network structure of blockchain.

[0064] Compliance and Auditing: The transparency and immutability of blockchain help banks meet compliance and auditing requirements. All transaction records are open and transparent, easy for regulators to review, and transaction records cannot be tampered with, ensuring the authenticity and integrity of audit data.

[0065] Smart contract application: Smart contracts can automatically execute and manage the payment process. The verification, clearing, and reconciliation of the payment process can be automatically executed by smart contracts, and payment and settlement can be automatically performed according to preset conditions;

[0066] Pre-conditions may include contract terms, time limits, etc.

[0067] However, unified management of payment channels based on centralized data processing has some disadvantages:

[0068] In terms of data management and transparency, transaction data is stored in a centralized database and managed by a central agency. Data transparency depends on the policies of the central agency. Usually only insiders and regulators can view complete transaction data, and data may be tampered with by internal or external attacks. Data integrity depends on the security measures of the central agency.

[0069] In terms of security, unified management of payment channels based on centralized data processing relies on traditional security measures such as firewalls and intrusion detection systems to protect data. If the central server is attacked or fails, the entire system may be affected, and it relies entirely on internal permission management mechanisms to prevent unauthorized access and operations.

[0070] In terms of efficiency and performance, traditional databases and payment systems may have targeted performance optimizations and be able to handle a large number of concurrent transactions, but the processing efficiency for real-time transactions is not ideal. When the system cannot meet the needs of business innovation, regulatory requirements and exception processing, manual intervention may be the last resort to meet customer requirements. However, due to the weaknesses of manual processing in terms of timeliness and cost, deploying operator resources and optimizing business processing processes will become a huge challenge, and manual execution of certain operations and rules may lead to delays and human errors.

[0071] In terms of compliance and regulation, KYC / AML checks are required to be performed manually, which is inefficient, data is stored in a centralized database, and audit transparency depends on the institution’s internal controls.

[0072] Distributed ledgers are the basic feature of blockchain technology. Distributed ledgers are used to record all transactions, ensuring that all data is open, transparent and cannot be tampered with, making transaction history traceable, helping to identify and track suspicious activities. All nodes hold the same copy of the ledger, ensuring that all parties can obtain the latest transaction data in real time, thereby achieving timely monitoring and response.

[0073] The identities of the transaction parties are confirmed through a decentralized identity verification system, and the transactions are verified using public key infrastructure and digital signature technology;

[0074] Users generate globally unique decentralized identifiers through the blockchain network. Decentralized identifiers can only be controlled by users and used for identification and judgment without the intervention of central registration agencies. At the same time, decentralized identifiers can be used across different systems and platforms. Decentralized identifiers can be used for identity authentication in multiple services, greatly simplifying the user experience;

[0075] Each user has a pair of public and private keys. The public key is used to publicly verify the identity, and the private key is used to sign transactions and access control to ensure the security and non-forgeability of identity information. The user uses the private key to sign the identity information, and the verifier uses the public key to verify the authenticity of the signature to ensure the source and integrity of the information.

[0076] Existing identity authentication methods mostly use multi-factor authentication, which verifies the user's identity by combining knowledge factors, possession factors, and intrinsic factors. Knowledge factors are information that users know, usually passwords, PIN codes, or answers to security questions. In the identity authentication of the transaction process, the unified management of payment channels requires the transaction parties to perform the first step of verification based on knowledge factors. However, when the transaction parties verify the knowledge factors, they may leak the knowledge factors used for verification. In the unified management of payment channels based on blockchain technology, the transaction parties have control over their own identity information and can use zero-knowledge proofs to show their identity information without leaking specific data, thereby ensuring data privacy and security during the verification process and avoiding the exposure of sensitive information. For example, users prove that they are legal residents without revealing specific addresses, and transaction parties can selectively disclose identity information. Through privacy sandbox technology, anonymous credentials are used to prove identity. Without exposing the user's identity, service providers can still ensure authenticity, thereby allowing the safe use and sharing of data and ensuring data privacy and security. Due to the existence of public key encryption and digital signatures, the transaction party's identity data is always secure and cannot be forged.

[0077] On the other hand, since the distributed ledger is decentralized and has no single control point, it can reduce the risk of single point failure and data leakage. The verification network is also decentralized and can jointly verify the identity through multiple independent nodes, which can also prevent single point failure and centralized attacks, and improve the robustness and reliability of the system.

[0078] The identity authentication system is integrated through the use of zero-knowledge proof libraries and tools. Multi-factor authentication includes three authentication factors: knowledge factors, possession factors, and intrinsic factors to enhance the security of identity authentication. Multiple independent verification nodes are deployed on the blockchain network to ensure decentralization and high reliability. The consensus mechanism is implemented using a consensus algorithm to ensure the security and consistency of the verification process.

[0079] Commonly used consensus algorithms include proof of work, proof of stake, Byzantine fault tolerance, etc.

[0080] Set up smart contracts to automatically execute transactions according to the set anti-fraud rules;

[0081] A smart contract is an automated script program that runs on the blockchain and can automatically execute contract terms when preset conditions are met. Once the logic of the smart contract is deployed on the blockchain, it can be automatically executed when specific conditions are met without the intervention of an intermediary or third party. The code of the smart contract is visible to all network participants, and anyone can verify the functions and behaviors of the contract. Smart contracts run in a distributed manner in the blockchain network and do not rely on any single central server or entity. Once a smart contract is deployed on the blockchain, the code and functions cannot be changed unless the contract itself allows modification. Due to the immutability and transparency of smart contracts, participants can trust that the contract will be executed according to the written logic without worrying about fraud or breach of contract. Smart contracts can write complex logic and rules to enable them to handle various types of transactions and agreements. With the development of blockchain technology, smart contracts can achieve interoperability across different blockchain networks, expanding the scope of application.

[0082] Example 2

[0083] In a typical embodiment, a smart contract is deployed in a unified management method for payment channels based on blockchain technology. When the smart contract is running on the blockchain, it can collect transaction data, including transaction amount, transaction frequency, transaction time, transaction party identity information, and transaction party geographic location. Rules are defined for the smart contract, and anti-fraud operations are performed using the smart contract. The rules are based on the following factors:

[0084] Detection is performed by setting thresholds, including amount thresholds and frequency thresholds;

[0085] Amount threshold: set the upper limit of a single transaction or the cumulative transaction amount within a period of time;

[0086] Frequency threshold: Set the maximum number of transactions allowed within a certain period of time;

[0087] Identification through transaction patterns, including geographic location and time;

[0088] Geographic location anomalies: detect whether the location where the transaction occurred is inconsistent with the user's usual location;

[0089] Time anomaly: detect whether the transaction occurs during the user's abnormal activity time period;

[0090] Analyze through transaction behavior, including historical comparison and device verification;

[0091] Historical data comparison: Compare with the user's historical transaction behavior to detect abnormal patterns;

[0092] Device information: Check if the transaction is from a commonly used device;

[0093] Filtering through blacklists and whitelists;

[0094] Blacklist: Block transactions with known fraudsters or suspicious accounts;

[0095] Whitelist: Allow transactions with trusted accounts to reduce false positives;

[0096] It should be noted that the blacklist is a list of prohibited or restricted transaction entities. These entities may be blacklisted by financial institutions, regulators or payment platforms themselves for various reasons. The blacklist can assist payment platforms in identifying and managing risks.

[0097] After the transaction is initiated, the transaction data is submitted to the blockchain and processed by the smart contract. The smart contract verifies the transaction according to the set anti-fraud rules and calculates the transaction fraud index by establishing a logistic regression model. The calculation expression is: In the formula, Th is the threshold rule coefficient, Pr is the pattern recognition coefficient, Ba is the behavior analysis coefficient, and Li is the blacklist filtering coefficient;

[0098] The calculation expression of the threshold rule coefficient is Th = α*Am+β*Fe, where Am is the transaction amount, Fe is the transaction frequency, α and β are the weight coefficients of the transaction amount and transaction frequency, and both α and β are greater than 0;

[0099] The calculation expression of the pattern recognition coefficient is Pr = γ*Ge + δ*Ti, where Ge is the geographical factor value, Ti is the time factor value, γ and δ are the weight coefficients of the geographical factor value and the time factor value respectively, and both γ and δ are greater than 0;

[0100] When the transaction location is inconsistent with the usual location of the transaction party, the geographical factor value is 0; when the transaction location is consistent with the usual location of the transaction party, the geographical factor value is 1;

[0101] When the transaction occurs outside the user's normal activity time period, the time factor value is 0; when the transaction occurs within the user's normal activity time period, the time factor value is 1;

[0102] The calculation expression of the behavior analysis coefficient is Ba=E*(Ab+1), where E is the equipment factor value. When the equipment used by the transaction party is not a common equipment, E takes a value of 1. When the equipment used by the transaction party is a common equipment, E takes a value of 0. Ab is the abnormal coefficient. The abnormal coefficient is a parameter for judging fraudulent transactions based on the machine learning model by analyzing and processing the historical transaction data of the transaction party through the use of a machine learning algorithm. When the machine learning model identifies the transaction behavior as a fraud risk transaction, Ab takes a value of 1. When the machine learning model identifies the transaction behavior as a stable and safe transaction, Ab takes a value of 0.

[0103] When the transaction party is included in the blacklist, the blacklist filtering coefficient is 0, and when the transaction party is not included in the blacklist, the blacklist filtering coefficient is 1.

[0104] Use machine learning algorithms to analyze transaction data, identify suspicious transactions, and automatically mark suspicious transactions;

[0105] The method for calculating the anomaly coefficient using machine learning algorithm is:

[0106] Collect historical transaction data of the transaction parties and classify the historical transaction data, including tendency data, account data, relationship data, and external data. The tendency data includes the user's online frequency, online path, IP address change frequency, and browsing behavior characteristics. The account data includes the user's account type, historical successful transactions, and historical failed transactions. The relationship data includes the transaction connection between the user's accounts and the behavior information of binding social accounts. The external data includes third-party credit scores, regional risk levels, and market volatility information.

[0107] Online access path includes the user's operating system type and online channel; browsing behavior characteristics include browsing interface, dwell time, and click behavior; user account types are divided into personal accounts and collective accounts;

[0108] Establish machine learning models based on trend data, account data, relationship data, and external data, clean the data, including processing missing values ​​and outliers, standardizing data of different scales, converting categorical data into numerical form, extracting features based on different data, using feature selection algorithms to select features, reducing dimensions to improve model performance, training data using supervised learning methods, optimizing model hyperparameters through grid search or random search, optimizing feature extraction and selection, and using trained models to judge transaction events;

[0109] Feature selection algorithms include chi-square test, mutual information, L1 regularization, etc. Commonly used supervised learning algorithms include decision tree, random forest, gradient boosting tree, support vector machine, neural network, etc.;

[0110] Setting a transaction fraud threshold and comparing the calculated transaction fraud index with the transaction fraud threshold;

[0111] When the abnormal coefficient of a transaction event is 1, and the transaction fraud index is greater than or equal to the transaction fraud threshold, the transaction event is marked as a crisis event, the transaction activity is frozen and reported to the management personnel;

[0112] When the abnormal coefficient of a transaction event is 1 and the transaction fraud index is less than the transaction fraud threshold, the transaction event is marked as a risk event, the transaction activity is monitored separately and the management personnel are informed;

[0113] When the abnormal coefficient of a transaction event is 0 and the transaction fraud index is greater than or equal to the transaction fraud threshold, the transaction event is marked as a critical event, the transaction activity is marked, and the management personnel are reminded to verify it;

[0114] When the abnormal coefficient of a transaction event is 0 and the transaction fraud index is less than the transaction fraud threshold, the transaction event is marked as a stable event and no marking operation is performed.

[0115] This application optimizes the risk management in the unified management of payment channels by using blockchain technology, integrates different payment channels using the distributed architecture of blockchain, records all transactions of payment channels in the same blockchain ledger, and uses the security and confidentiality of blockchain to provide personalized security services for transaction users. It defines constraints on the blockchain's anti-fraud smart contracts based on machine learning algorithms, combines the transparent and automated execution of smart contracts, avoids errors and omissions in manual operations, and further reduces the risk management costs of unified payments.

[0116] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0117] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of computer program goods. The computer program goods include one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center by wired or wireless (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0118] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0119] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0120] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the system described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0121] If the functions are implemented in the form of software functional units and sold or used as independent goods, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can essentially or in other words, the part that contributes to the prior art or the part of the technical solution can be embodied in the form of software goods, and the computer software goods are stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, and other media that can store program codes.

[0122] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A unified payment channel management method, characterized in that: The specific steps include: Build a distributed ledger based on blockchain technology and share real-time data among all nodes in the distributed ledger; The identities of the transaction parties are confirmed through a decentralized identity verification system, and the transactions are verified using public key infrastructure and digital signature technology; Set up smart contracts to automatically execute transactions according to the set anti-fraud rules; Combined with machine learning algorithms, abnormal transactions are identified and classified by analyzing transaction data, and abnormal transactions are automatically marked; How to set up a smart contract: When running on the blockchain, smart contracts can collect transaction data, including transaction amount, transaction frequency, transaction time, transaction party identity information, and transaction party geographic location. Rules are defined for smart contracts and anti-fraud operations are performed using smart contracts. The rules are based on the following factors: Detection is performed by setting thresholds, including amount thresholds and frequency thresholds; Amount threshold: set the upper limit of a single transaction or the cumulative transaction amount within a period of time; Frequency threshold: Set the maximum number of transactions allowed within a certain period of time; Identification through transaction patterns, including geographic location and time; Geographic location anomalies: detect whether the location where the transaction occurred is inconsistent with the user's usual location; Time anomaly: detect whether the transaction occurs during the user's abnormal activity time period; Analyze through transaction behavior, including historical comparison and device verification; Historical data comparison: Compare with the user's historical transaction behavior to detect abnormal patterns; Device information: Check if the transaction is from a commonly used device; Filtering through blacklists and whitelists; Blacklist: Block transactions with known fraudsters or suspicious accounts; Whitelist: Allow transactions with trusted accounts; Methods for verifying transactions based on smart contracts: After the transaction is initiated, the transaction data is submitted to the blockchain and processed by the smart contract. The smart contract verifies the transaction according to the set anti-fraud rules and calculates the transaction fraud index by establishing a logistic regression model. The calculation expression is: In the formula, Th is the threshold rule coefficient, Pr is the pattern recognition coefficient, Ba is the behavior analysis coefficient, and Li is the blacklist filtering coefficient; When the transaction party is included in the blacklist, the blacklist filtering coefficient is 0, and when the transaction party is not included in the blacklist, the blacklist filtering coefficient is 1; The calculation method of threshold rule coefficient and pattern recognition coefficient is: The calculation expression of the threshold rule coefficient is Th = α*Am+β*Fe, where Am is the transaction amount, Fe is the transaction frequency, α and β are the weight coefficients of the transaction amount and transaction frequency, and both α and β are greater than 0; The calculation expression of the pattern recognition coefficient is Pr = γ*Ge + δ*Ti, where Ge is the geographical factor value, Ti is the time factor value, γ and δ are the weight coefficients of the geographical factor value and the time factor value respectively, and both γ and δ are greater than 0; When the transaction location is inconsistent with the usual location of the transaction party, the geographical factor value is 0; when the transaction location is consistent with the usual location of the transaction party, the geographical factor value is 1; When the transaction occurs outside the user's normal activity time period, the time factor value is 0; when the transaction occurs within the user's normal activity time period, the time factor value is 1.

2. A payment channel unified management method according to claim 1, characterized in that: The logic for obtaining the behavior analysis coefficient is: The calculation expression of the behavior analysis coefficient is Ba=E*(Ab+1), where E is the equipment factor value. When the equipment used by the transaction party is not a common equipment, E takes a value of 1. When the equipment used by the transaction party is a common equipment, E takes a value of 0. Ab is the abnormality coefficient. The abnormality coefficient is a parameter for judging fraudulent transactions based on the machine learning model by analyzing and processing the historical transaction data of the transaction party using a machine learning algorithm. When the machine learning model identifies the transaction behavior as a fraud risk transaction, Ab takes a value of 1. When the machine learning model identifies the transaction behavior as a stable and safe transaction, Ab takes a value of 0.

3. A payment channel unified management method according to claim 2, characterized in that: The logic of calculating the anomaly coefficient using machine learning methods is: Collect historical transaction data of the transaction parties and classify the historical transaction data, including tendency data, account data, relationship data, and external data. The tendency data includes the user's online frequency, online path, IP address change frequency, and browsing behavior characteristics. The account data includes the user's account type, historical successful transactions, and historical failed transactions. The relationship data includes the transaction connection between the user's accounts and the behavior information of binding social accounts. The external data includes third-party credit scores, regional risk levels, and market volatility information. Online access path includes the user's operating system type and online channel; browsing behavior characteristics include browsing interface, dwell time, and click behavior; user account types are divided into personal accounts and collective accounts; Establish machine learning models based on tendency data, account data, relationship data, and external data, clean the data, including processing missing values ​​and outliers, standardizing data of different scales, converting categorical data into numerical form, extracting features based on different data, using feature selection algorithms to select features, reducing dimensions to improve model performance, training data using supervised learning methods, optimizing model hyperparameters through grid search or random search, optimizing feature extraction and selection, and using trained models to judge transaction events.

4. A payment channel unified management method according to claim 3, characterized in that: Combined with machine learning algorithms, the method for identifying and classifying abnormal transactions and automatically marking abnormal transactions is as follows: Setting a transaction fraud threshold and comparing the calculated transaction fraud index with the transaction fraud threshold; When the abnormal coefficient of a transaction event is 1, and the transaction fraud index is greater than or equal to the transaction fraud threshold, the transaction event is marked as a crisis event, the transaction activity is frozen and reported to the management personnel; When the abnormal coefficient of a transaction event is 1 and the transaction fraud index is less than the transaction fraud threshold, the transaction event is marked as a risk event, the transaction activity is monitored separately and the management personnel are informed; When the abnormal coefficient of a transaction event is 0 and the transaction fraud index is greater than or equal to the transaction fraud threshold, the transaction event is marked as a critical event, the transaction activity is marked, and the management personnel are reminded to verify it; When the abnormal coefficient of a transaction event is 0 and the transaction fraud index is less than the transaction fraud threshold, the transaction event is marked as a stable event and no marking operation is performed.