Consumption financial relation graph system

Abstract

The invention discloses a consumer financial relation graph system, and belongs to the technical field of data processing and financial science and technology. The system constructs an association network between people and between people and key attributes based on a graph database technology, and solves the problems of insufficient relationship mining, limited risk control capability and low data processing efficiency in existing consumption financial data management. The system comprises a data source module, a data processing module, a data storage module, a feature processing module, a data analysis module, an application service module and a data visualization module. Through data cleaning, integration and feature extraction, the system can efficiently store and analyze consumer financial data, realizes risk assessment, fraud detection and credit scoring by using a machine learning algorithm, and visually displays an analysis result through a visual map. The system supports real-time calculation of hundreds of billions of nodes and edges, meets the requirements of complex business scenes, and remarkably improves the risk control capability, the precision marketing level and the data processing efficiency in the consumer finance field.

Description

Technical Field

[0001] The present invention relates to the technical field of financial data processing and analysis, and in particular to a consumer finance relationship graph system. Background Art

[0002] With the rapid development of the consumer finance industry, financial institutions have accumulated massive amounts of customer data. However, existing technologies have many limitations in data management and application. Traditional data storage mainly relies on structured databases. Although they have certain advantages in storing and managing relational data, they are unable to cope with complex relational networks and dynamically associated data. In addition, existing technologies have the following problems in mining and utilizing the potential value of relational data:

[0003] Relational data mining is not enough

[0004] Currently, financial institutions focus more on the storage and analysis of structured data, and less on the relationship network between users and the potential connections between people and key attributes. This limitation makes it impossible to fully identify customer behavior patterns, potential needs and associated risks.

[0005] Limited risk control capabilities

[0006] Financial business scenarios are becoming increasingly complex, and fraudulent activities such as false applications, gang fraud, and intermediary packaging are emerging in an endless stream. However, traditional risk control systems have weak monitoring capabilities for these complex connections and are often unable to achieve real-time warnings and rapid responses, causing financial institutions to face high losses.

[0007] Inefficient data management

[0008] The rapid growth of massive data has put higher demands on existing data storage and processing technologies. Traditional systems are unable to meet the needs of efficient storage, real-time query and complex analysis when faced with large-scale, dynamically updated relational data.

[0009] Insufficient precision marketing

[0010] Existing technologies lack in-depth analysis of user relationships and behavioral characteristics, and are unable to accurately understand customer needs and develop personalized marketing strategies, resulting in low customer conversion rates.

[0011] In response to the above problems, the introduction of relationship graph technology provides a new solution for efficient management and in-depth analysis of consumer finance data. Through the architectural design based on the graph database, the relationship graph system can build a complex relationship network, sort out the implicit connections between customers and between customers and attributes, and thus achieve the following goals:

[0012] Reveal implicit relationship networks: By analyzing massive data, we build association maps between people and between people and attributes to discover potential customer needs.

[0013] Enhance risk control capabilities: By identifying abnormal behavioral characteristics in customer relationships, high-risk behaviors such as false applications and group fraud can be effectively discovered.

[0014] Optimize customer management: Based on the analysis results of the relationship map, it helps financial institutions to achieve functions such as customer precision marketing, repair lost customer information and reverse case investigation.

[0015] Improve system performance: Design a highly scalable and highly available graph database architecture to support real-time feature calculation and rule extraction on billions of nodes and edges to meet the growing business needs of financial institutions.

[0016] In summary, in response to the shortcomings of existing technologies in the field of consumer finance, the present invention proposes a consumer finance data management and analysis system based on a relationship graph, aiming to solve the pain points of current technologies in terms of efficiency, risk control, marketing, etc., and provide efficient and intelligent technical support for the consumer finance industry. Summary of the invention

[0017] The present invention proposes a consumer finance relationship graph system, which focuses on big data analysis and application in the field of consumer finance, and is committed to building a relationship network between people and between people and attributes through the introduction of graph database technology, so as to solve the shortcomings of existing technologies in data management, association analysis and risk control.

[0018] System Structure

[0019] The present invention designs a highly available and highly scalable relationship graph system architecture, which mainly includes multiple sub-modules such as data source module, data processing module, data storage module, feature processing module, data analysis module, application service module and data visualization module. The system supports real-time and offline analysis, and can meet the data processing and second-level response requirements of billions of nodes and edges.

[0020] Data source module

[0021] This module is used to access data from various sources, including file systems, databases, and sensors, and provide raw data input. Through the system interface, it supports the reading of static data such as log files, as well as the access of dynamic data streams, providing a variety of input channels for subsequent data processing.

[0022] Data processing module

[0023] The data processing module standardizes multi-source data through steps such as cleaning queues, data integration, and data verification to form a clear and consistent data flow. Specifically, it includes:

[0024] Data cleaning queue: automatically filter invalid, redundant or erroneous data;

[0025] Data verification module: ensure data consistency and reliability;

[0026] Data integration module: merges data from multiple sources into a unified format according to set rules to support subsequent processing and storage.

[0027] Data storage module

[0028] The data storage module combines high-performance graph databases (such as Neo4j) and search engines (such as ElasticSearch) to ensure efficient data storage and flexible queries. Graph databases are used to store complex relational data, while ElasticSearch supports full-text indexing and quickly completes fuzzy queries and associated searches. In addition, the module also combines traditional relational databases to support hybrid data requirements.

[0029] Feature Processing Module

[0030] This module extracts features from the processed standardized data and processes them into vector form suitable for modeling and analysis. The feature extraction rules can be dynamically adjusted according to business needs, supporting custom feature generation in specific scenarios.

[0031] Data Analysis Module

[0032] The data analysis module is the core of the entire system, which implements in-depth analysis through data mining and machine learning algorithms. The module contains multiple sub-functions:

[0033] Risk assessment: Analyze user behavior trajectories and identify potential risks;

[0034] Fraud detection: monitor data abnormal patterns in real time to detect fraudulent behavior;

[0035] Credit scoring: Generate accurate credit scores for users based on historical behavior and data models.

[0036] Application service module

[0037] The application service module provides decision support and business strategy suggestions for external systems through load balancing, interface services and other functions. Service support includes real-time anti-fraud warning, precision marketing strategy optimization, lost customer information restoration, case tracing and other scenarios.

[0038] Data Visualization Module

[0039] The data visualization module visualizes complex analysis results and supports users to view data associations in the form of charts, relationship maps, etc. The module functions include:

[0040] Legend filtering: helps users quickly locate specific nodes and relationships;

[0041] Fuzzy matching query: improve query flexibility;

[0042] Historical trajectory analysis: Displays user behavior trajectories to provide support for risk control.

[0043] System architecture and technical implementation

[0044] The system architecture is based on distributed design and adopts a multi-layer modular structure, including load layer, service layer, middleware layer and data layer. The specific implementation is as follows:

[0045] Load layer: Through Tengine and its cache module, external requests are processed and distributed to backend services.

[0046] Service layer: Use the SpringBoot framework to build service interfaces and provide efficient business logic processing.

[0047] Middleware layer: Introduce components such as Kafka message queue, Redis cache, ZooKeeper coordination service, etc. to ensure the high availability and distributed coordination capabilities of the system.

[0048] Data layer: Combines data storage technologies such as Hbase, ElasticSearch and Neo4j to meet the needs of high-frequency queries, full-text retrieval and relational storage.

[0049] Beneficial Effects

[0050] The present invention significantly improves the data processing and risk control capabilities in the field of consumer finance through the innovative application of relationship graph technology. The main technical effects include:

[0051] Improve data processing efficiency: Through modular architecture and distributed storage, intelligent processing of large-scale data can be achieved.

[0052] Reduce financial risks: Accurately identify fraudulent activities and reduce losses caused by false applications and group fraud.

[0053] Optimize customer management: Identify potential customer needs through relationship network analysis to achieve precision marketing and post-loan management.

[0054] Enhance system scalability: Adopt a highly available distributed architecture to support business growth and complex analysis needs.

[0055] In summary, the present invention provides a comprehensive and efficient consumer finance relationship map system, which solves the data management and analysis problems in the prior art and provides financial institutions with powerful decision-making support and risk control capabilities. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only for the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0057] Figure 1 A schematic diagram of the logical structure of an embodiment of the present invention;

[0058] Figure 2 A functional architecture diagram of an embodiment of the present invention;

[0059] Figure 3 A schematic diagram of the technical architecture of an embodiment of the present invention;

[0060] Figure 4 Schematic diagram of network topology according to an embodiment of the present invention. DETAILED DESCRIPTION

[0061] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. At the same time, it is explained here that in order to make the embodiments more detailed, the following embodiments are the best and preferred embodiments, and those skilled in the art may also adopt other alternatives to implement some known technologies; and the accompanying drawings are only for more specific description of the embodiments, and are not intended to specifically limit the present invention.

[0062] It should be noted that the references to "one embodiment", "an embodiment", "an exemplary embodiment", "some embodiments" and the like in the specification indicate that the embodiments described may include specific features, structures or characteristics, but not every embodiment may include the specific features, structures or characteristics. In addition, when a specific feature, structure or characteristic is described in conjunction with an embodiment, it should be within the knowledge of a person skilled in the art to implement such feature, structure or characteristic in conjunction with other embodiments (whether or not explicitly described).

[0063] In general, a term can be understood, at least in part, from its use in context. For example, depending, at least in part, on the context, the term "one or more" as used herein can be used to describe any feature, structure, or characteristic in the singular sense, or can be used to describe a combination of features, structures, or characteristics in the plural sense. Additionally, the term "based on" can be understood as not necessarily intended to convey an exclusive set of factors, but can instead, depending, at least in part, on the context, allow for the presence of other factors that are not necessarily explicitly described.

[0064] See also Figures 1 to 4

[0065] 1. Preparation for implementation

[0066] Before implementing this consumer finance relationship map system, you need to ensure that you have the corresponding hardware equipment, including but not limited to servers, storage devices, etc., and the server should have sufficient computing power and storage capacity to meet the system operation requirements. At the same time, prepare the required software environment, such as installing relevant operating systems, database management systems, middleware, etc., and ensure the compatibility between various software.

[0067] 2. Logical Structure Implementation

[0068] (I) Data source module

[0069] Data is collected from multiple channels such as the application system, credit review system, collection system, and external data. For example, the application system provides original data such as personal information, application amount, and application time filled in by customers when applying for loans; the credit review system contributes data related to customers' credit assessment; and the collection system provides data such as customers' repayment status. These data are accurately transmitted to the data source module through data interfaces or data transmission protocols.

[0070] The collected data is preliminarily sorted and formatted to meet the requirements of subsequent processing. For example, text data in different formats is uniformly converted into a specific structured format to ensure data consistency and availability.

[0071] (II) Data processing module

[0072] The data cleaning queue cleans the original data to remove duplicate data, erroneous data, incomplete data, etc. For example, by setting data format verification rules, data that does not meet the requirements can be screened out and corrected or deleted.

[0073] The data verification and integration module verifies the cleaned data to ensure the accuracy and completeness of the data. At the same time, it integrates data from different data sources, establishes the relationship between the data, and forms a standardized data format for storage and analysis. For example, the data generated by customers in different business links are associated and integrated through the customer's unique identifier.

[0074] (III) Data storage module

[0075] Store the processed data in a relational database (such as Oracle, etc.), and use the database's transaction processing, data consistency maintenance and other functions to ensure the safe storage of data. According to the type and purpose of the data, reasonably design the database table structure to improve data storage and query efficiency.

[0076] For some data that needs to be processed in real time or frequently queried, you can use caching technology (such as Redis) to cache it in memory to speed up data access and improve system performance.

[0077] (IV) Application Service Module

[0078] Load balancing components (such as Tengine) reasonably distribute requests from external systems according to the server load to ensure balanced utilization of server resources and avoid single point overload. For example, by monitoring the server's CPU usage, memory usage and other indicators, the request distribution strategy can be dynamically adjusted.

[0079] Node processing and relationship processing services further process data to meet the needs of different business scenarios. For example, according to specific business rules, data is aggregated, calculated, and other operations are performed to generate intermediate data or final results that can be used for analysis and decision-making.

[0080] Through interface services, data access and interactive interfaces are provided for external systems (such as other financial business systems, data analysis platforms, etc.) to achieve data sharing and collaborative work. Interface services should follow unified interface specifications to ensure the accuracy and security of data transmission.

[0081] 3. Functional Architecture Implementation

[0082] (I) Data Visualization Module

[0083] According to user needs, relevant data is obtained from the database or cache and data set preparation is performed. For example, if a user needs to view the distribution of customer loan amounts within a certain period of time, the data visualization module extracts the corresponding data from the data storage layer and performs pre-processing operations such as data cleaning and conversion.

[0084] Use visualization libraries (such as Echarts, etc.) or related tools to select appropriate chart types (such as bar charts, line charts, pie charts, etc.) according to data characteristics to make charts. Present the made charts to users in an intuitive way, such as displaying them on the user interface, or generating reports for users to download and view.

[0085] (II) Feature Processing Module

[0086] Extract valuable features for consumer finance analysis from the original data, such as customer consumption behavior characteristics, credit characteristics, etc. For example, by analyzing the customer's historical consumption records, extract features such as consumption frequency, average consumption amount, and consumption category distribution.

[0087] The extracted features are cleaned and transformed to make them suitable for subsequent data analysis and model training. For example, the features are normalized to convert data of different dimensions to the same order of magnitude to improve the accuracy and stability of the model.

[0088] (III) Data Analysis Module

[0089] Implement various data analysis algorithms, such as overlapping community detection algorithms to discover potential relationships between customer groups, unsupervised clustering algorithms to classify customers, label propagation algorithms to predict customer attributes, etc. Select appropriate algorithms to analyze and process data based on business needs.

[0090] Use data analysis results to support financial decision-making, such as identifying potential fraud risk customer groups, predicting customer default probability, etc. Through in-depth analysis of customer behavior data, provide a basis for financial institutions to formulate accurate marketing strategies, risk control strategies, etc.

[0091] (IV) Other functions

[0092] Graph query function: Users can quickly query relevant customer information and their relationships in the relationship graph by entering specific conditions (such as customer name, ID number, loan product type, etc.). The system performs efficient searches in graph databases (such as Neo4j, etc.) based on the query conditions entered by the user, and displays the query results to the user in an intuitive way, such as presenting the association paths and relationship strengths between customers in the form of a graphical graph.

[0093] Anti-fraud function: By real-time monitoring of customer transaction data, behavior data, etc., combined with preset fraud detection rules and models, potential fraud behaviors can be discovered in a timely manner. For example, when a customer's transaction behavior shows abnormal fluctuations (such as frequent large-value transactions in a short period of time, abnormal transaction locations, etc.), the system automatically triggers the early warning mechanism and notifies relevant personnel to verify and handle.

[0094] Post-loan management function: Real-time tracking and management of customers after loan issuance, including monitoring customer repayments, and timely initiating collection processes when customers are overdue; analyzing trends in customer repayment behavior, assessing changes in customer credit risk, and providing reference for financial institutions to adjust their loan strategies.

[0095] 4. Technical Architecture Implementation

[0096] (I) Load layer

[0097] Configure Tengine as a Web server and reverse proxy server to optimize server performance. Set cache strategies, such as reasonably determining the cache validity period and cache content based on the access frequency and timeliness of data, to increase data reading speed and reduce the pressure on the backend server.

[0098] KeepaLive is used to achieve high availability of the server. Through heartbeat detection and other mechanisms, the server operation status is monitored in real time. When the main server fails, it automatically switches to the backup server to ensure the continuous and stable operation of the system.

[0099] (II) Service Layer

[0100] Use the SpringBoot framework to quickly build application services and simplify the development process. Through its convention-based feature, it reduces the tedious work of writing configuration files and improves development efficiency. At the same time, use SpringBoot's built-in components (such as SpringSecurity, etc.) to enhance the security of the application, such as user authentication and authorization management.

[0101] With the help of WebService and RestfulControl, interface services under different protocols are provided to meet the diverse external system access requirements. According to business requirements, reasonable interface definitions and interface documents are designed to ensure the ease of use and scalability of the interface. For example, Restful interfaces are provided to third-party partners to facilitate them to obtain data such as customer credit assessment results.

[0102] (III) Middleware

[0103] Deploy Redis as an application cache to store frequently accessed data or temporary calculation results. Set the data structure of Redis appropriately.

[0104] (IV) Data Layer

[0105] Use Hbase to store large-scale structured data, and design a reasonable table structure and storage mode based on the characteristics of the data and business needs. For example, store basic customer information and transaction records in different tables, and establish appropriate indexes to improve data storage and query efficiency.

[0106] Use ElasticSearch to achieve powerful full-text search and analysis functions. Create indexes for data that require full-text search (such as customer notes, contract terms, etc.) so that users can quickly and accurately search for relevant information. At the same time, use its aggregation analysis function to perform statistical analysis on the data, such as counting the number of loan applications and repayment amount distribution in different time periods.

[0107] With the help of hive, data warehouse related operations are supported, and complex SQL queries are supported. By writing appropriate SQL statements, data can be extracted from the data warehouse for analysis and report generation. For example, customer risk assessment reports, business operation reports, etc. can be generated regularly to provide data support for the decision-making of financial institutions.

[0108] Use Kettle for data integration and ETL operations to extract, transform and load data between different data sources. For example, extract data from external data sources (such as third-party credit rating agency data), clean and transform it, and then load it into the data storage of this system to ensure data integrity and consistency.

[0109] (V) Message Queuing and Coordination Services

[0110] Deploy Kafka message queues to achieve asynchronous data transmission and high-throughput processing. Data is transmitted between the data source module and the data processing module through Kafka to ensure reliable data transmission and avoid data loss or duplicate processing. For example, when the application system generates new customer application data, the data is sent to the Kafka queue, and the data processing module obtains the data from the queue for subsequent processing.

[0111] ZooKeeper cluster is used to maintain the system status and configuration information to achieve coordination and synchronization between nodes. For example, in a server cluster, ZooKeeper is used for service discovery and distributed lock management to ensure the coordination between system components.

[0112] 5. Network topology diagram implementation

[0113] 1. Server deployment

[0114] Configure the main database server to store core data and ensure its high availability and data security. Use data backup, redundant storage and other technologies to prevent data loss. For example, regularly perform full and incremental backups of the database and store them in an offsite backup center.

[0115] Deploy partitioned database servers to partition and store data according to business needs to improve data query and processing efficiency. For example, partition data by customer region, business type, etc., so that specific areas or business-related data can be quickly located and obtained when querying.

[0116] Install the relationship graph business server, run the relationship graph related services and applications, and realize the data association analysis and graph construction functions. This server is connected to other servers through a high-speed network to ensure the timeliness and accuracy of data transmission.

[0117] 2. Service Collaboration

[0118] The distribution server receives requests from external systems and distributes them to appropriate application servers based on the load balancing strategy. For example, based on the current load of the application server, response time and other indicators, the optimal server is dynamically selected to handle the request, improving the overall performance and response speed of the system.

[0119] The application server processes specific business logic, interacts with the data storage layer and other servers, completes tasks such as data query, analysis, and processing, and returns the results to the distribution server or directly to the external system. For example, when a user queries a customer relationship graph, the application server obtains relevant data from the database, generates graph data through a relationship graph construction algorithm, and converts it into a visual format and returns it to the user.

[0120] (III) External system interaction

[0121] Data interaction and integration with other financial business systems (such as core business systems, risk management systems, etc.) are carried out through interfaces. For example, basic customer information and transaction data are obtained from the core business system, and risk assessment results are fed back to the risk management system to achieve business process collaboration and data sharing.

[0122] Provide data access interfaces for external partners (such as third-party data providers, regulators, etc.) to ensure secure data sharing and compliant use. In terms of interface design, adopt security authentication, authorization management and other measures to protect data privacy and system security. For example, third-party data providers provide supplementary data to the system through authorized interfaces, and regulators can obtain necessary business data for regulatory review.

[0123] Through the above specific implementation methods, this consumer finance relationship map system can effectively integrate various data resources, realize intelligent data processing and analysis, provide financial institutions with comprehensive risk management, precision marketing, decision support and other functions, and enhance the competitiveness and operational efficiency of financial institutions. During the implementation process, reasonable configuration and optimization should be carried out according to actual business needs and technical environment to ensure the stable operation and sustainable development of the system. At the same time, pay close attention to the development of industry technology, and upgrade and improve the system in a timely manner to adapt to the ever-changing market needs and regulatory requirements.

[0124] The present invention covers any substitution, modification, equivalent method and scheme made on the essence and scope of the present invention. In order to make the public have a thorough understanding of the present invention, specific details are described in detail in the following preferred embodiments of the present invention, but those skilled in the art can fully understand the present invention without the description of these details. In addition, in order to avoid unnecessary confusion about the essence of the present invention, well-known methods, processes, procedures, components and circuits are not described in detail.

[0125] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A consumer finance relationship graph system, characterized in that: Includes the following modules: Data source module: used to collect data from file systems, databases and sensors, and provide raw data for the data processing module; Data processing module: including data cleaning, data integration and data verification functions, standardizing the original data; Data storage module: uses graph database and relational database to store processed data and support fast query; Feature processing module: extracts feature vectors from standardized data and further processes them to support analysis and modeling; Data analysis module: Modeling, analysis and risk assessment of consumer finance data through data mining and machine learning algorithms; Application service module: provides interface services based on analysis results, supporting functions such as precision marketing, anti-fraud and post-loan management; Data visualization module: visualize data analysis results in the form of charts, graphs, etc. to facilitate users' intuitive understanding.

2. The consumer finance relationship graph system according to claim 1, characterized in that: The data source module includes: File system interface, used to receive log files and externally input static data; Database interface, used to obtain real-time updated data from relational database; Sensor interface for acquiring dynamically generated data streams.

3. The consumer finance relationship graph system according to claim 1 or 2, characterized in that: The data processing module further comprises: Data cleaning queue, used to automatically filter invalid or redundant data; Data verification module, used to check the consistency and accuracy of data; Data integration module, used to merge data from multiple sources into a unified format.

4. The consumer finance relationship graph system according to claim 1, characterized in that: The data storage module uses a graph database (Neo4j) to store and retrieve relational data, and combines it with ElasticSearch for fast full-text indexing and searching.

5. The consumer finance relationship graph system according to claim 1, characterized in that: The feature processing module supports custom feature extraction rules and implements dynamic adjustment through scripts to adapt to diverse business needs.

6. The consumer finance relationship graph system according to claim 1, characterized in that: The data analysis module comprises: Risk assessment submodule: identifies potential risks by analyzing user behavior trajectories; Fraud detection submodule: real-time monitoring of abnormal patterns in data; Credit score submodule: Calculates credit scores based on the user's historical behavior.

7. The consumer finance relationship graph system according to claim 1, characterized in that: The application service module further includes: Decision engine, used to generate business strategies based on analysis results; The interface service layer supports external systems to call relationship graph analysis results in the form of APIs.

8. The consumer finance relationship graph system according to claim 1, characterized in that: The data visualization module supports legend screening, fuzzy query and historical track display functions, which are used to quickly locate key nodes and their associations.

9. The consumer finance relationship graph system according to claim 1, characterized in that: The system adopts a distributed architecture, combined with load balancing and message queue technology, to ensure stability and scalability in a high-concurrency environment.

Images