Banking system multi-dimensional splitting optimization method and configurable processing platform

Abstract

The invention discloses a multidimensional splitting optimization method and a configurable processing platform for a bank system, and the method comprises the steps: longitudinally elaborating an overall splitting formation mode of a new bank business system, and taking a loan accounting business line as a background, and proposing a testability design which should be possessed by a bank loan accounting system and a configuration design of bank loan accounting. According to the method, the system performance, flexibility and expandability are remarkably improved, the data integrity and safety are guaranteed, and the upgrading of a banking business system is powerfully promoted.

Description

Technical Field

[0001] The present invention relates to the financial industry, and in particular to a multi-dimensional splitting optimization method and a configurable processing platform for a banking system. Background Art

[0002] With economic development and advancements in Internet technology, more and more people are choosing to conduct banking services online, such as online transfers and payments. However, most banks currently still use traditional centralized systems to handle various banking services. This system has many shortcomings, such as:

[0003] 1) Poor maintainability

[0004] High code complexity: As business functionality continues to grow, the codebase of a monolithic architecture becomes increasingly large, intermingling code from different business modules, reducing code readability and comprehensibility. For example, in a monolithic application encompassing multiple banking functions, code for credit cards, loans, and savings services might all be concentrated in a single project. This makes it difficult for developers to quickly locate the relevant code when modifying or extending a particular function, and can easily introduce errors.

[0005] Difficulties in deployment and updating: Monolithic applications typically require packaging the entire project into a single executable file or deployment package. Updates to small features require redeploying the entire application, which not only increases deployment time and costs but can also cause system failures during deployment, impacting normal business operations.

[0006] 2) Limited scalability

[0007] Inflexible resource allocation: In a monolithic architecture, all business functions run in the same process space and share the same server resources. When the load on certain business functions increases, it's difficult to scale resources for them individually. The only option is to upgrade the entire server or increase the number of servers. However, this can cause other business functions to occupy excessive resources, resulting in resource waste. For example, during a bank's peak transaction period, the concurrent volume of transfer transactions increases significantly. However, due to the limitations of the monolithic architecture, resource expansion for just the transfer function is impossible; the entire system must be scaled, increasing costs.

[0008] Insufficient horizontal scalability: Monolithic applications are typically designed and optimized for a single server, making horizontal scalability—that is, adding servers to share the load—impossible. When business volume grows to a certain level, the performance bottleneck of a single server becomes a major constraint on system growth. Transforming a monolithic application into a distributed one often requires a large-scale restructuring of the entire system, which is costly.

[0009] 3) Inflexible technology selection

[0010] Single technology stack: Monolithic architectures typically select a fixed technology stack at the beginning of a project. As the business evolves, new business requirements may arise that require different technologies. However, due to the limitations of a monolithic architecture, it is difficult to introduce new technologies without disrupting the entire system. For example, when a bank needs to introduce big data analytics to process massive amounts of customer transaction data, the traditional databases and development frameworks used in a monolithic architecture do not support big data processing. Implementing this functionality requires a complete system overhaul, increasing the difficulty of technology upgrades.

[0011] Difficulty adapting to new technological developments: With the continuous advancement of technology, new development frameworks, database technologies, and cloud computing platforms are constantly emerging. Due to its holistic and complex nature, monolithic architectures struggle to keep pace with new technological developments, putting banks at a disadvantage in terms of technological innovation and competitiveness.

[0012] 4) High reliability and stability risks

[0013] Single point of failure has a significant impact: In a monolithic architecture, all business functions rely on the same application and server. If a failure in that application or server, such as a hardware failure, software vulnerability, or network issue, occurs, the entire system will not function properly, impacting all banking operations. For example, if a bank's core business system uses a monolithic architecture, a server failure could prevent customers from withdrawing, transferring, or making deposits, resulting in significant losses for both the bank and its customers.

[0014] Difficulty troubleshooting and recovery: Due to the tight coupling between business functions in a monolithic architecture, it is difficult to quickly locate the specific cause and location of a system failure. Furthermore, recovery can take a long time because the entire system needs to be inspected and repaired, resulting in prolonged business interruption.

[0015] To overcome these shortcomings and meet the demands of modern banking, many financial institutions are considering transitioning from traditional centralized systems to distributed systems. Distributed systems network multiple small computers together to accomplish specific tasks. These systems offer the following advantages:

[0016] 1. Addressing Business Complexity: With the continuous expansion of banking services, they have encompassed diverse areas such as retail banking, corporate finance, and investment banking. Each business segment encompasses numerous complex functions and processes. For example, retail banking's credit card business, from application and approval, credit limit management, to repayment and installment payments, involves a vast array of business rules and data processing. Using a microservices architecture, these complex business functions can be broken down into multiple independent microservices, each focused on a specific business function. This clarifies business logic and facilitates development, maintenance, and management.

[0017] 2. Meeting high concurrency requirements: In certain business scenarios, such as online and mobile banking, banks may experience large numbers of concurrent users accessing their services during peak periods. For example, during shopping festivals like Singles' Day (Singles' Day), banks' payment systems must process massive volumes of transaction requests. A microservices architecture allows for flexible and independent scaling of microservices based on the workload of different businesses. This allows for increased server resources to handle high concurrency, ensuring system stability and responsiveness, and preventing system crashes caused by single points of failure or insufficient resources.

[0018] 3. Improve business agility: Banking businesses must rapidly respond to market changes and customer needs. Traditional monolithic architectures often require time-consuming redeployment and testing of the entire system when upgrading business functions or expanding new businesses. Microservices architectures, on the other hand, allow for independent development, testing, and deployment of each microservice. New businesses or features can be quickly added to the system as microservices, significantly shortening the launch cycle and improving business agility. This allows banks to quickly launch new products and services in the face of fierce market competition and meet the personalized needs of their customers.

[0019] 4. Facilitates technological upgrades and innovation: The banking industry needs to continuously keep pace with the times and introduce new technologies and tools to improve service quality and efficiency. A microservices architecture allows banks to select the most appropriate technology stack based on the characteristics and needs of each microservice. For example, microservices for data processing and analysis can utilize big data processing technologies and distributed databases; while transaction microservices with high real-time requirements can utilize high-performance in-memory databases. Furthermore, each microservice can independently undergo technological upgrades and innovations without impacting the normal operation of other services, facilitating the gradual evolution and upgrade of banks' technological architectures.

[0020] 5. Enhanced system reliability and stability: In a microservices architecture, each microservice is independent and independent. A failure in a microservice affects only that service, without paralyzing the entire system. Banks can use monitoring and self-healing mechanisms to promptly detect and address faulty microservices, ensuring overall system reliability. Furthermore, communication between microservices utilizes lightweight protocols, such as RESTful APIs. This loosely coupled architecture enhances system stability and allows for better response to emergencies and abnormal events.

[0021] 6. Compliance with regulatory requirements: The banking industry is subject to strict regulatory requirements, such as data security, privacy protection, and business continuity. A microservices architecture can better meet these requirements by isolating different business functions and controlling permissions. Each microservice can implement appropriate security measures and data protection mechanisms based on its business nature and data sensitivity, ensuring the security of customer information and bank data. Furthermore, the scalability and high availability of a microservices architecture help banks meet regulatory requirements for business continuity and ensure the stable operation of financial services.

[0022] However, in practical applications, how to smoothly transition to a distributed system without changing the existing system architecture is a challenge. This paper proposes a banking system optimization and transformation solution based on a microservice architecture to address this problem. Summary of the Invention

[0023] In view of this, the present application provides a multi-dimensional splitting optimization method and a configurable processing platform for a banking system to address the defects in the prior art.

[0024] The technical solution of the embodiment of the present application is as follows: A multi-dimensional splitting optimization method for a banking system comprises the following steps:

[0025] (a) Data object domain dimension splitting: The customer information module is separated into a customer master data system (CIF), which includes basic customer information management, credit authorization verification, and customer relationship map construction functions; the contract information module is separated into a credit contract management system (CMS), which implements contract text storage, contract status tracking, and intelligent clause analysis;

[0026] (b) Data volume separation: High-frequency transaction flow data is separated into a separate Loan Transaction System (LTS), which uses a time-series database to store transaction details and supports second-level queries on billions of data points. Core accounting functions are separated into a separate Loan Account System (LAS), which provides account lifecycle management and credit limit control.

[0027] (c) Capability platform integration: File management, distributed cache, and message queue components are abstracted into a unified front-end platform, providing infrastructure capability support for each microservice through standardized interfaces;

[0028] (d) Distributed feature design: Provide a read-only interface for the CIF system, configure an event-driven mechanism for the CMS system, implement idempotent operations based on globally unique serial numbers for the LAS system, and design a time series index optimization strategy for the LTS system.

[0029] Furthermore, the read-only interface implementation of the customer master data system (CIF) includes:

[0030] An API gateway is used to control the routing of query requests to the loan system, a read-write separation database architecture is used to ensure data consistency, and a request frequency circuit breaker mechanism is set to prevent service overload.

[0031] Furthermore, the event-driven mechanism of the credit contract management system (CMS) specifically includes:

[0032] When a contract is signed or its terms are changed, a standardized Doma in Event event is published through the message middleware. The subscription system triggers subsequent business processes based on the event type. The event content includes the contract version hash value and the change timestamp.

[0033] It also includes a banking system design method that supports multi-mode testing, including:

[0034] (a) Time decoupling mechanism: Configure an independent system clock module to support the separation of business processing time and natural time, allowing timestamp injection to simulate business scenarios at any time point during testing;

[0035] (b) Branch control switch: A function switch matrix is built into the interest calculation module, and the interest calculation strategy, interest settlement period, and fault tolerance rules are dynamically adjusted through the configuration center;

[0036] (c) Transaction integrity assurance: Generate a globally unique tracking ID for each transaction, implement idempotent verification based on the serial number hash value in the anti-duplicate subsystem, and establish an abnormal transaction recovery channel to support data rollback and compensation operations.

[0037] Furthermore, the time decoupling mechanism further includes:

[0038] A logical clock synchronization algorithm is used in a distributed environment, and cross-system time consistency is ensured through the NTP server deviation compensation mechanism. In test mode, the virtual clock offset can be set at the transaction granularity.

[0039] A configurable processing platform for a banking system, comprising:

[0040] (a) Multi-dimensional trigger engine: Integrates timed triggers, event listeners, and condition monitors, and supports configuring complex trigger strategies through combined rules;

[0041] (b) Dynamic rule loading module: uses hot deployment technology to achieve visual configuration and real-time effectiveness of interest calculation rules and fee calculation models;

[0042] (c) Batch task scheduling platform: Provides a graphical workflow designer that supports automatic selection of serial / parallel execution mode based on data volume, and configures breakpoint resumption and exception warning strategies.

[0043] Furthermore, it also includes:

[0044] The multi-dimensional trigger engine includes the following configuration modes:

[0045] Scheduled trigger mode: uses enhanced cron expressions to define interest payment tasks, supporting differentiated configurations for natural days, working days, and holidays;

[0046] Event-driven mode: monitor loan success events through the message bus to trigger subsequent interest accumulation and account processing pipelines;

[0047] Conditional trigger mode: monitor account balance changes in real time and automatically trigger the interest recalculation process when early repayment is detected.

[0048] Furthermore, the implementation of the dynamic rule loading module includes:

[0049] The interest calculation algorithm is abstracted into a rule decision tree, with each decision node associated with a business parameter weight. Grayscale release is achieved through rule version management, and a historical rule snapshot backtracking function is provided.

[0050] The beneficial effects of the technical solutions provided in the embodiments of the present application include at least:

[0051] This application discloses a method for optimizing and transforming a banking business system and a corresponding banking business system. By decomposing an existing banking business system and then optimizing each of the decomposed business modules, a more efficient, secure, and easy-to-maintain distributed system is ultimately achieved. This method can also help financial institutions reduce operating costs, improve service quality, and better respond to market changes and customer needs. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0053] Figure 1 This application presents a multi-dimensional splitting optimization method for a banking system and a schematic diagram of a configurable processing platform. Figure 1 .

[0054] Figure 2 This application presents a multi-dimensional splitting optimization method for a banking system and a schematic diagram of a configurable processing platform. Figure 2 .

[0055] Figure 3 This application provides a loan issuance process diagram of a multi-dimensional splitting optimization method for a banking system and a configurable processing platform. DETAILED DESCRIPTION

[0056] Exemplary embodiments will be described in detail herein, with examples illustrated in the accompanying drawings. In the following description, when referring to the drawings, identical numerals in different figures represent identical or similar elements, unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all embodiments consistent with the present application. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present application, as detailed in the appended claims.

[0057] 1. Technical solution and principle

[0058] (1) Method of splitting banking business system

[0059] The present invention proposes a multi-dimensional splitting optimization method for a banking system, and constructs a reasonable microservice ecosystem through a three-dimensional splitting strategy. First, based on the data object domain dimension, customer information, account information, product information, etc. are split into independent management centers. For example, the customer master data system (CIF) is responsible for managing basic customer information, credit authorization and relationship maps, and provides a read-only interface for loan system queries to avoid data redundancy; the credit contract management system (CMS) focuses on contract text storage, contract status tracking and clause parsing, and adopts an event-driven model to publish DomainEvent when signing or changing. Secondly, based on the data magnitude dimension, the flow registration and account management functions are separated into independent systems. For example, the loan transaction flow system (LTS) uses a time series database to store transaction details such as loan issuance, repayment, and extension, realizing query of billions of data in seconds. Finally, starting from the capability platform dimension, file management, cache management, and message queues are integrated into a unified front-end system.

[0060] This approach is applicable to a variety of scenarios, including evolving from a traditional monolithic architecture to a microservices architecture, expanding the service ecosystem of an existing microservices architecture, and designing and creating microservices from scratch. Its innovation lies in establishing a standard for the formation of microservices and classifying them based on this standard. It also defines common considerations and restrictions for different categories, facilitating subsequent targeted resource expansion supervision, security governance, data management, and log management for microservices.

[0061]

[0062]

[0063] (2) Testability Design of Banking Business Systems

[0064] The present invention also relates to a testability design scheme for a banking system. Testability is crucial to improving the maintainability and stability of the system. Taking the loan interest calculation process as an example, it is necessary to verify the correctness of the data at each time point and each time it is processed. Therefore, the testability of the system is particularly critical. Unlike traditional testing methods, this design emphasizes that in the process of testing and developing code, the support capabilities for test conditions should be fully considered. For example, it solves the problem of inconsistency between the system operation date and the natural time; sets flexible switch configurations for functional branches; enhances the anti-duplicate mechanism of transactions, the support capability for secondary processing of abnormal data, and the anti-transaction function.

[0065] The innovation of this invention is that it emphasizes the necessity of testability in business processing algorithms, requires that the development process fully considers the support for testing work, and fundamentally improves the test feasibility and effectiveness of the system.

[0066] (3) Configuration design of banking business system

[0067] This paper proposes a configuration-based design for a banking system, aiming to improve system testability. Configuration management is primarily considered from three perspectives: first, test case configuration, which supports recording and reusing data from each stage to facilitate testing operations; second, test template configuration, which enables batch task activation and filtering of processed data; and third, the periodic operation of batch data, which supports not only daily data processing but also monthly, annual, and other periodic batch processing requirements.

[0068] In terms of business scenario trigger types, it covers timed triggering (such as mortgage interest payment on the 20th of each month, implemented through cron expressions, task scheduling system and interest calculation rule engine), event triggering (interest accrues immediately after successful loan disbursement, and transaction events are monitored and triggered asynchronously), balance change triggering (interest is adjusted after early repayment, based on transaction flow monitoring and real-time calculation), cycle switching triggering (interest calculation period is changed after loan extension, and rule updates are driven by contract change events) and manual triggering (interest recalculation required by regulators, achieved through buttons on the backend management interface combined with double-signature approval).

[0069]

[0070]

[0071] The innovation of this invention lies in emphasizing the necessity of configurability in business processing algorithms. By proposing the above three methods to improve testability, the flexibility and scalability of system configuration are achieved, providing strong support for the testing and optimization of business systems.

[0072] 2. Implementation Methods

[0073] Banking business system architecture construction

[0074] Data Object Management Center Cluster: Build a data object management center cluster that includes customer information management nodes, account management nodes, and product information nodes. Server resources should be rationally allocated based on data volume and access frequency. For example, for a customer information management node with a large number of customers and frequent queries, a distributed database architecture can be adopted to store customer information data across multiple server nodes to improve data read and write speeds. By constructing a data association graph, a data synchronization and association mechanism is established between nodes to ensure data consistency. For example, when customer information changes, the associated account and product information is automatically updated through the data synchronization mechanism.

[0075] Business Function Service Cluster: Build a business function service cluster that includes a transaction registration system, an account management system, and a risk control system. The transaction registration system receives transaction data from various business modules, generates and stores a unique transaction number for the entire district. The account management system receives accounting rules and transaction amount information through a standardized accounting interface and performs accounting processing and registration. The risk control system conducts risk assessment and monitoring of business transactions. For example, in a transfer transaction, the transaction registration system records the transfer transaction information, the account management system performs accounting processing based on the transfer amount and accounting rules, and the risk control system monitors the transfer in real time to determine if there are any risks.

[0076] Technical Support Platform: A technical support platform was established, encompassing a standardized front-end service gateway, an intermediate table management engine, and a scheduled task scheduler. The standardized front-end service gateway integrates atomic service interfaces for file transfer, cache management, and message queuing, providing unified technical services for other systems. The intermediate table management engine implemented a keyword tagging strategy, automatically generating temporary data views with a "view" suffix for batch task processing. The scheduled task scheduler, comprising an interest calculation module, a task association module, and an exception monitoring module, is responsible for scheduling and managing loan interest-earning scheduled tasks.

[0077] Business process implementation

[0078] Loan disbursement process: When a loan is disbursed, a request is first sent to the loan accounting system. After disbursement, the loan accounting system records the data in the debit note table. It then calls the settlement system to process the transfer and the account flow registration system to register the loan disbursement. The account flow system asynchronously aggregates the transactions to the general ledger system. Throughout this process, data exchange between various systems is achieved through standardized interfaces, ensuring smooth business processes. For example, data is transferred between the loan accounting system and the account flow registration system using standardized accounting interfaces based on accounting rules and transaction amounts, ensuring accurate account registration.

[0079] Loan interest generation process: After the loan is issued, the loan interest generation timed task system is started through the timed task scheduler. The interest calculation module calculates the accumulated interest according to the set scheduling frequency (such as minute level), and supports the segmented cumulative calculation of normal interest and penalty interest. When the accumulated interest is completed, the task association module automatically triggers the interest settlement operation. The exception monitoring module monitors the abnormal situation in the interest calculation process in real time. Once an abnormality occurs, the exception handling rollback mechanism is immediately activated to ensure data consistency. For example, if the system power is cut off during the interest calculation process, the exception monitoring module detects the calculation interruption, and the exception handling rollback mechanism rolls back the calculated part of the interest data and recalculates after the system returns to normal.

[0080] System configuration and parameter settings

[0081] Scheduling Parameter Settings: In the scheduled task scheduler, set the scheduling frequency in minutes, hours, days, months, or years, and the scheduling span in intervals of 1-60 minutes, 1-24 hours, 1-31 days, or 1-12 months. Flexible scheduling parameters can be configured based on the needs of different loan businesses. For example, for ultra-short-term loans, you can set the scheduling frequency to minutes and the scheduling span to 15 minutes to achieve more granular interest accumulation.

[0082] Intermediate table management parameter settings: In the intermediate table management engine, set the keyword suffix to "view" to distinguish the intermediate data of different batch processing tasks. Also, set the storage period and cleanup rules for the intermediate tables based on the characteristics of the batch tasks. For example, for daily batch accounting processing tasks, set the intermediate tables to automatically clean up the day after the task is completed to free up system storage space.

[0083] Through the above specific implementation methods, the banking business system optimization and transformation method proposed in the present invention can be fully realized, and a high-performance, flexible, secure and stable banking business system can be constructed to meet the needs of modern financial business development.

[0084] Obviously, those skilled in the art may make various changes and modifications to this application without departing from the spirit and scope of this application. Thus, if these modifications and variations of this application fall within the scope of the claims of this application and their equivalents, this application is intended to include these modifications and variations.

Claims

1. A multi-dimensional splitting optimization method for a banking system, characterized by: The following steps are involved: (a) Data object domain dimension splitting: The customer information module is separated into a customer master data system (CIF), which includes basic customer information management, credit authorization verification, and customer relationship map construction functions; the contract information module is separated into a credit contract management system (CMS), which implements contract text storage, contract status tracking, and intelligent clause analysis; (b) Data volume separation: High-frequency transaction flow data is separated into a separate Loan Transaction System (LTS), which uses a time-series database to store transaction details and supports second-level queries on billions of data points. Core accounting functions are separated into a separate Loan Account System (LAS), which provides account lifecycle management and credit limit control. (c) Capability platform integration: File management, distributed cache, and message queue components are abstracted into a unified front-end platform, providing infrastructure capability support for each microservice through standardized interfaces; (d) Distributed feature design: Provide a read-only interface for the CIF system, configure an event-driven mechanism for the CMS system, implement idempotent operations based on globally unique serial numbers for the LAS system, and design a time series index optimization strategy for the LTS system.

2. The method according to claim 1, characterized in that The read-only interface implementation method of the customer master data system (CIF) includes: An API gateway is used to control the routing of query requests to the loan system, a read-write separation database architecture is used to ensure data consistency, and a request frequency circuit breaker mechanism is set to prevent service overload.

3. The method according to claim 1, characterized in that The event-driven mechanism of the credit contract management system (CMS) specifically includes: When a contract is signed or its terms are changed, a standardized Doma in Event event is published through the message middleware. The subscription system triggers subsequent business processes based on the event type. The event content includes the contract version hash value and the change timestamp.

4. A multi-dimensional splitting optimization method for a banking system, characterized in that: It also includes a banking system design method that supports multi-mode testing, including: (a) Time decoupling mechanism: Configure an independent system clock module to support the separation of business processing time and natural time, allowing timestamp injection to simulate business scenarios at any time point during testing; (b) Branch control switch: A function switch matrix is built into the interest calculation module, and the interest calculation strategy, interest settlement period, and fault tolerance rules are dynamically adjusted through the configuration center; (c) Transaction integrity assurance: Generate a globally unique tracking ID for each transaction, implement idempotent verification based on the serial number hash value in the anti-duplicate subsystem, and establish an abnormal transaction recovery channel to support data rollback and compensation operations.

5. The method according to claim 4, characterized in that The time decoupling mechanism further includes: A logical clock synchronization algorithm is used in a distributed environment, and cross-system time consistency is ensured through the NTP server deviation compensation mechanism. In test mode, the virtual clock offset can be set at the transaction granularity.

6. A configurable processing platform for a banking system, characterized in that: include: (a) Multi-dimensional trigger engine: Integrates timed triggers, event listeners, and condition monitors, and supports configuring complex trigger strategies through combined rules; (b) Dynamic rule loading module: uses hot deployment technology to achieve visual configuration and real-time effectiveness of interest calculation rules and fee calculation models; (c) Batch task scheduling platform: Provides a graphical workflow designer that supports automatic selection of serial / parallel execution mode based on data volume, and configures breakpoint resumption and exception warning strategies.

7. The configurable processing platform according to claim 6, wherein: include: The multi-dimensional trigger engine includes the following configuration modes: Scheduled trigger mode: uses enhanced cron expressions to define interest payment tasks, supporting differentiated configurations for natural days, working days, and holidays; Event-driven mode: monitor loan success events through the message bus to trigger subsequent interest accumulation and account processing pipelines; Conditional trigger mode: monitor account balance changes in real time and automatically trigger the interest recalculation process when early repayment is detected.

8. The configurable processing platform according to claim 6, wherein: The implementation of the dynamic rule loading module includes: The interest calculation algorithm is abstracted into a rule decision tree, with each decision node associated with a business parameter weight. Grayscale release is achieved through rule version management, and a historical rule snapshot backtracking function is provided.

Images