An online big data intelligent risk control system for the banking industry

By designing an online big data intelligent risk control system and using a risk control model with multi-dimensional data and self-learning ability, the problem of single and inobjective risk types of existing risk control systems is solved, and more efficient and accurate risks are achieved Evaluation and management.

CN118037422BActive Publication Date: 2024-11-22SHENZHEN XINTAI DIGITAL TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410128232.9
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-01-30
Publication Date
2024-11-22
Estimated Expiration
2044-01-30

AI Technical Summary

Technical Problem

The existing risk control system has a single type of risk control and lacks objectivity, resulting in low accuracy of risk control results and affecting the use effect.

Method used

Design an online big data intelligent risk control system for the banking industry, including user information collection module, user classification module, user monitoring module and data processing module. Through real-time collection, classification and processing of user information, risk control is generated Warning information, using multi-dimensional data and risk control models of self-learning ability for risk assessment and monitoring.

Benefits of technology

It improves the accuracy and efficiency of risk assessment, can identify risks more objectively and impartially, reduce non-performing loans and recovery expenses, optimize risk management processes, and improve the degree of automation and the accuracy of risk warning.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118037422B_ABST
    Figure CN118037422B_ABST
Patent Text Reader

Abstract

The present invention discloses an online big data intelligent risk control system for the banking industry, comprising a user information collection module, a user classification module, a user monitoring module, a data processing module and an information sending module; the user information collection module is used for collecting user information to obtain user information, and the user information includes user available amount information, user historical consumption times, historical consumption amount, historical consumption location and historical risk control times; the user classification module is used for processing user information to obtain user classification information, and the user classification information includes one type of user, two types of users, three types of users and four types of users; the user monitoring module is used for real-time monitoring of users according to the user classification information to obtain real-time user monitoring information; the present invention can better generate risk control warnings and more intelligently and wisely perform online risk control of banking business.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of risk control systems, and in particular to an online big data intelligent risk control system for the banking industry. Background Art

[0002] Risk control refers to the various measures and methods taken by risk managers to eliminate or reduce the possibility of risk events, or the risk controllers to reduce the losses caused by risk events. Bank risk control refers to bank risk management and control, which is an indispensable part of banking operations. It involves the process of identifying, evaluating, monitoring and controlling potential risks, aiming to protect the bank's assets and capital while ensuring the sound development of its business;

[0003] During the bank risk control process, the risk control system will be applied to ensure the safety of funds.

[0004] In actual use, the existing risk control system has a single risk control type and is not objective enough, resulting in low accuracy of risk control results, which has a certain impact on the use of the risk control system. Therefore, an online big data intelligent risk control system for the banking industry has been extracted. Summary of the invention

[0005] In view of the defects in the prior art, the present invention provides an online big data intelligent risk control system for the banking industry, including a user information collection module, a user classification module, a user monitoring module, a data processing module and an information sending module;

[0006] The user information collection module is used to collect user information to obtain user information, including user available amount information, user historical consumption times, historical consumption amount, historical consumption location and historical risk control times;

[0007] The user classification module is used to process the user information to obtain user classification information, wherein the user classification information includes first-class users, second-class users, third-class users and fourth-class users;

[0008] The user monitoring module is used to monitor users in real time according to user classification information to obtain real-time user monitoring information;

[0009] The data processing module is used to process real-time user monitoring information and user classification information to generate user risk control warning information;

[0010] The information sending module is used to send the user risk control warning information to a preset receiving terminal after the information is generated.

[0011] Furthermore, the specific process of the user information collection module collecting user information is as follows:

[0012] Determine the objectives and scope of data collection and clarify the data that needs to be collected;

[0013] Select the corresponding data collection method according to the data source and type;

[0014] Determine the frequency and time of data collection based on business needs and data change frequency.

[0015] Perform data cleaning and preprocessing, including removing duplicate data, filling missing values ​​and processing outliers;

[0016] Data storage and management: the collected data is stored in a preset data storage device and managed and maintained. The management and maintenance content includes data backup, data archiving, and data security protection.

[0017] Furthermore, the specific process of user classification by the user classification module is as follows: extracting the collected user information, and extracting the user's available amount information, the user's historical consumption times, the historical consumption amount, the historical consumption location and the historical risk control times from the user information;

[0018] The user's available balance information is marked as G1, the user's historical consumption times are marked as G2, the historical consumption amount is marked as G3, and the historical risk control times are marked as G4;

[0019] Process the historical consumption location information to obtain location evaluation parameters; mark it as G5;

[0020] The ratio of the historical consumption amount G3 to the user's historical consumption times G2 is calculated to obtain the single consumption evaluation parameter Gg1;

[0021] When the historical risk control times G4 is less than or equal to the preset value q1, the user's available amount information G1 is greater than the preset value a1, the single consumption evaluation parameter Gg1 is greater than the preset value b1, the user's historical consumption times G2 is greater than the preset value c1, and the location evaluation parameter G5 is greater than the preset value d1, the user is assigned to the second category of users;

[0022] When the historical risk control times G4 is less than or equal to the preset value q1, the user's available amount information is greater than the preset value a1, the single consumption evaluation parameter Gg1 is between the preset values ​​b1 and b2, the user's historical consumption times G2 is greater than the preset value c1, and the location evaluation parameter G5 is less than or equal to the preset value d1, the user is assigned to the third category;

[0023] When the historical risk control times G4 is less than or equal to the preset value q1, the user's available amount information is between the preset values ​​a1 and a2, the single consumption evaluation parameter Gg1 is greater than the preset value b1, the user's historical consumption times G2 is greater than the preset value c1, and the location evaluation parameter G5 is less than or equal to the preset value d1, the user is assigned to the third category;

[0024] When the historical risk control times G4 is greater than the preset value q1, and the user's available amount information G1 is greater than the preset value a1, the user is directly assigned to Category 1 without considering other factors;

[0025] When the historical risk control times G4 is greater than the preset value q1, and the user's available amount information is between the preset values ​​a1 and a2, that is, without considering other factors, the user is directly assigned to the second category;

[0026] In addition to the above process, the remaining users are divided into four categories of users.

[0027] Furthermore, the specific processing procedure for processing the historical consumption location information to obtain the location evaluation parameter is as follows: extract the historical consumption location, which is the consumption location information of the x times closest to the current time, import the x times of consumption location information into a preset map software, collect the distance information between the x times of consumption locations, extract the number of distance information whose distance is greater than a preset value in the distance information between the x times of consumption locations, and obtain the location evaluation parameter.

[0028] Furthermore, the user monitoring module is used to monitor the user in real time according to the user classification information to obtain the real-time user monitoring information in the following specific process:

[0029] Extract user classification information. When the user classification information is a type of user, that is, the user monitoring information is collected every time the user consumes;

[0030] When the user classification information is a Class II user, the user monitoring information is collected after the user consumes at least e1 times;

[0031] When the user classification information is the third category of users, the user monitoring information is collected after the user's consumption times reaches e2;

[0032] When the user classification information is four types of users, the user monitoring information is collected every preset time period.

[0033] Furthermore, the specific processing process of the data processing module for processing the real-time user monitoring information and user classification information to generate user risk control warning information is as follows: extract the collected user classification information, model the user classification information and import it into a preset risk control model library, determine the grading type of the standard risk control model for user classification, then obtain the real-time user model after modeling the real-time user monitoring information, match the real-time user model with the corresponding grading type, and when a standard risk control model with a similarity greater than a preset value to the real-time user model is matched, user risk control warning information is generated.

[0034] Furthermore, the risk control model library presets a first-class risk control model, a second-class risk control module, a third-class risk control module and a fourth-class risk control module;

[0035] One type of risk control model corresponds to one type of user, the second type of risk control module corresponds to the second type of user, the third type of risk control module corresponds to the third type of user, and the fourth type of risk control module corresponds to the fourth type of user.

[0036] Furthermore, the specific process of obtaining the real-time user model after modeling the real-time user monitoring information is as follows: extracting the collected real-time user monitoring information, the real-time user monitoring information includes consumption time, consumption location, consumption amount, and consumption frequency, and then extracting features from the real-time user monitoring information, and extracting indicators reflecting the characteristics of user consumption behavior through in-depth analysis of the user's consumption data, including consumption frequency, average consumption amount, and consumption location distribution. Then, using the extracted features, the bank constructs a variety of risk control models, including a real-time logistic regression model, a real-time decision tree model, and a real-time random forest model;

[0037] In the process of matching the real-time user model with the corresponding classification type, the established real-time logistic regression model, real-time decision tree model and real-time random forest model are synchronized and matched with the models in the preset risk control model library.

[0038] Furthermore, after the real-time logistic regression model, the real-time decision tree model and the real-time random forest model are established, it is also necessary to perform model evaluation and optimization on the real-time logistic regression model, the real-time decision tree model and the real-time random forest model.

[0039] Furthermore, the determination process for model evaluation and optimization of the real-time logistic regression model, the real-time decision tree model and the real-time random forest model is as follows: the accuracy, recall rate and F1 value of the real-time logistic regression model, the real-time decision tree model and the real-time random forest model are collected. When the accuracy, recall rate and F1 value do not meet the preset conditions, it means that the model evaluation result is not ideal and the model needs to be adjusted and optimized.

[0040] The beneficial effects of the present invention are embodied in:

[0041] The present invention can improve the accuracy and efficiency of risk assessment. It can analyze the credit risk of users based on user behavior data. Through the real-time collection and analysis of these data, banks can make decisions faster to reduce risks and improve the accuracy and efficiency of risk assessment. Through the analysis of big data of user consumption behavior, banks can predict customer behavior and risks, thereby reducing non-performing loans and recovery costs, reducing costs and risks, and improving the degree of automation: the data collection and computing capabilities of big data can help banks establish real-time and efficient risk management views. With the help of multi-dimensional data, self-learning risk control models, and real-time calculation results, financial institutions can improve risk assessment and management capabilities, improve the degree of automation, and be more objective and fair: models built based on algorithms are more objective and fair. The combination of full data and subject behavior analysis can fully understand customer behavior and help financial institutions objectively and accurately identify risks and provide risk control prompts. Banks can conduct big data risk control through user consumption behavior, which can more comprehensively and accurately assess user credit risks, improve the accuracy and efficiency of risk assessment, reduce costs and risks, optimize risk management processes, increase the degree of automation, and more objectively and fairly identify risks and provide risk control prompts. In addition, the accuracy and speed of risk warnings are further improved through user classification and early warning settings, making the system more worthy of promotion and use. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following is a brief introduction to the drawings required for the specific embodiments or the description of the prior art. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn according to the actual scale.

[0043] Figure 1 It is a system block diagram of the present invention. DETAILED DESCRIPTION

[0044] The following embodiments of the technical solution of the present invention are described in detail in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and are therefore only used as examples, and cannot be used to limit the protection scope of the present invention.

[0045] It should be noted that, unless otherwise specified, the technical terms or scientific terms used in this application should have the common meanings understood by those skilled in the art to which the invention belongs.

[0046] like Figure 1 As shown, an online big data intelligent risk control system for the banking industry includes a user information collection module, a user classification module, a user monitoring module, a data processing module and an information sending module;

[0047] The user information collection module is used to collect user information to obtain user information, including user available amount information, user historical consumption times, historical consumption amount, historical consumption location and historical risk control times;

[0048] The user classification module is used to process the user information to obtain user classification information, wherein the user classification information includes first-class users, second-class users, third-class users and fourth-class users;

[0049] The user monitoring module is used to monitor users in real time according to user classification information to obtain real-time user monitoring information;

[0050] The data processing module is used to process real-time user monitoring information and user classification information to generate user risk control warning information;

[0051] The information sending module is used to send the user risk control warning information to a preset receiving terminal after the information is generated;

[0052] Through the real-time collection and analysis of these data, banks can make decisions faster to reduce risks and improve the accuracy and efficiency of risk assessment. Through the analysis of big data of user consumption behavior, banks can predict customer behavior and risks, thereby reducing non-performing loans and recovery costs, reducing costs and risks, and improving the degree of automation: the data collection and computing capabilities of big data can help banks establish a real-time and efficient risk management view. With the help of multi-dimensional data, self-learning risk control models, and real-time calculation results, financial institutions can improve risk assessment and control capabilities, improve the degree of automation, and be more objective and fair: models built based on algorithms are more objective and fair. Full data plus subject behavior analysis can fully understand customer behavior, help financial institutions objectively and accurately identify risks and provide risk control prompts. Banks use big data risk control through user consumption behavior to more comprehensively and accurately assess user credit risks, improve the accuracy and efficiency of risk assessment, reduce costs and risks, optimize risk management processes, improve the degree of automation, and more objectively and fairly identify risks and provide risk control prompts. In addition, the accuracy and speed of risk warnings are further improved through user classification and early warning settings, making the system more worthy of promotion and use.

[0053] The specific process of user information collection by the user information collection module is as follows:

[0054] Determine the objectives and scope of data collection and clarify the data that needs to be collected;

[0055] Select the corresponding data collection method according to the data source and type;

[0056] Determine the frequency and time of data collection based on business needs and data change frequency.

[0057] Perform data cleaning and preprocessing, including removing duplicate data, filling missing values ​​and processing outliers;

[0058] Data storage and management: storing the collected data in a preset data storage device and managing and maintaining it. The management and maintenance content includes data backup, data archiving, and data security protection;

[0059] The above process improves data quality and accuracy: through data cleaning and integration, duplicate, erroneous or incomplete data can be removed to ensure data accuracy and credibility, and improve data quality. Optimize resource allocation: through data collection and processing, we can better understand user needs and behaviors, thereby optimizing resource allocation, improving resource utilization efficiency and enterprise operating efficiency. Strengthen risk control capabilities: data collection and processing can provide more accurate and comprehensive data support for risk management, thereby better assessing risks and formulating risk control strategies, and improving the risk control capabilities of enterprises.

[0060] The specific process of user classification by the user classification module is as follows: extracting the collected user information, and extracting the user's available amount information, the user's historical consumption times, the historical consumption amount, the historical consumption location and the historical risk control times from the user information;

[0061] The user's available balance information is marked as G1, the user's historical consumption times are marked as G2, the historical consumption amount is marked as G3, and the historical risk control times are marked as G4;

[0062] Process the historical consumption location information to obtain location evaluation parameters; mark it as G5;

[0063] The ratio of the historical consumption amount G3 to the user's historical consumption times G2 is calculated to obtain the single consumption evaluation parameter Gg1;

[0064] When the historical risk control times G4 is less than or equal to the preset value q1, the user's available amount information G1 is greater than the preset value a1, the single consumption evaluation parameter Gg1 is greater than the preset value b1, the user's historical consumption times G2 is greater than the preset value c1, and the location evaluation parameter G5 is greater than the preset value d1, the user is assigned to the second category of users;

[0065] When the historical risk control times G4 is less than or equal to the preset value q1, the user's available amount information is greater than the preset value a1, the single consumption evaluation parameter Gg1 is between the preset values ​​b1 and b2, the user's historical consumption times G2 is greater than the preset value c1, and the location evaluation parameter G5 is less than or equal to the preset value d1, the user is assigned to the third category;

[0066] When the historical risk control times G4 is less than or equal to the preset value q1, the user's available amount information is between the preset values ​​a1 and a2, the single consumption evaluation parameter Gg1 is greater than the preset value b1, the user's historical consumption times G2 is greater than the preset value c1, and the location evaluation parameter G5 is less than or equal to the preset value d1, the user is assigned to the third category;

[0067] When the historical risk control times G4 is greater than the preset value q1, and the user's available amount information G1 is greater than the preset value a1, the user is directly assigned to Category 1 without considering other factors;

[0068] When the historical risk control times G4 is greater than the preset value q1, and the user's available amount information is between the preset values ​​a1 and a2, that is, without considering other factors, the user is directly assigned to the second category;

[0069] In addition to the above process, the remaining users are assigned to four categories of users;

[0070] a1>a2, b1>b2.

[0071] The specific process of processing the historical consumption location information to obtain the location evaluation parameter is as follows: extract the historical consumption location, which is the consumption location information of the x times closest to the current time, import the x times of consumption location information into the preset map software, collect the distance information between the x times of consumption locations, extract the number of distance information between the x times of consumption locations whose distance is greater than the preset value, that is, obtain the location evaluation parameter, x≥10;

[0072] Through the above process, users are classified in detail, and different types of risk control monitoring can be better performed for different types of users, thereby better ensuring the security of user funds.

[0073] The specific process of the user monitoring module for real-time monitoring of users according to user classification information to obtain real-time user monitoring information is as follows:

[0074] Extract user classification information. When the user classification information is a type of user, that is, the user monitoring information is collected every time the user consumes;

[0075] When the user classification information is a Class II user, the user monitoring information is collected after the user consumes at least e1 times;

[0076] When the user classification information is the third category of users, the user monitoring information is collected after the user's consumption times reaches e2;

[0077] When the user classification information is four types of users, the user monitoring information is collected at a preset time interval;

[0078] e1>e2, e1 and e2 are both positive integers, and the specific values ​​of e1 and e2 are manually set by the system administrator according to the actual situation.

[0079] The specific processing process of the data processing module processing the real-time user monitoring information and the user classification information to generate the user risk control warning information is as follows: extracting the collected user classification information, modeling the user classification information and importing it into a preset risk control model library, determining the classification type of the standard risk control model for user classification, then obtaining the real-time user model after modeling the real-time user monitoring information, matching the real-time user model with the corresponding classification type, and generating the user risk control warning information when a standard risk control model with a similarity greater than a preset value to the real-time user model is matched;

[0080] In the process of banks conducting big data risk control through user consumption behavior, classifying users can improve the accuracy and efficiency of risk assessment: by classifying users, banks can divide users into different risk levels and credit levels, and adopt different risk control strategies and credit policies for users of different levels, thereby improving the accuracy and efficiency of risk assessment.

[0081] Reduce costs and risks: By classifying users, banks can have a clearer understanding of users' credit status and risk levels, thereby formulating more scientific and effective risk management strategies and reducing costs and risks.

[0082] Improve user experience: By classifying users, banks can provide more personalized services and offers, improve user satisfaction and loyalty, and increase user stickiness.

[0083] Optimize risk management process: By classifying users, banks can adopt different risk control strategies for different categories of users, optimize risk management process and improve risk management efficiency.

[0084] Enhance risk warning capabilities: By classifying users, banks can more accurately predict users’ risk status, take corresponding risk control measures in advance, and enhance risk warning capabilities.

[0085] By classifying users during the process of big data risk control based on their consumption behavior, banks can more accurately assess user credit risks, improve the accuracy and efficiency of risk assessment, reduce costs and risks, optimize risk management processes, improve user experience, and enhance risk warning capabilities.

[0086] The risk control model library presets a first-class risk control model, a second-class risk control module, a third-class risk control module and a fourth-class risk control module;

[0087] One type of risk control model corresponds to one type of user, the second type of risk control module corresponds to the second type of user, the third type of risk control module corresponds to the third type of user, and the fourth type of risk control module corresponds to the fourth type of user.

[0088] The specific process of obtaining the real-time user model after modeling the real-time user monitoring information is as follows: extracting the collected real-time user monitoring information, the real-time user monitoring information includes consumption time, consumption location, consumption amount, and consumption frequency, and then extracting features from the real-time user monitoring information. By conducting an in-depth analysis of the user's consumption data, extracting indicators that reflect the characteristics of the user's consumption behavior, including consumption frequency, average consumption amount, and consumption location distribution, and then using the extracted features, the bank constructs a variety of risk control models, including a real-time logistic regression model, a real-time decision tree model, and a real-time random forest model;

[0089] In the process of matching the real-time user model with the corresponding classification type, the established real-time logistic regression model, real-time decision tree model and real-time random forest model are synchronized and matched with the models in the preset risk control model library.

[0090] After the real-time logistic regression model, the real-time decision tree model and the real-time random forest model are established, it is also necessary to perform model evaluation and optimization on the real-time logistic regression model, the real-time decision tree model and the real-time random forest model.

[0091] The determination process of model evaluation and optimization of the real-time logistic regression model, the real-time decision tree model and the real-time random forest model is as follows: the accuracy, recall rate and F1 value of the real-time logistic regression model, the real-time decision tree model and the real-time random forest model are collected. When the accuracy, recall rate and F1 value do not meet the preset conditions, it means that the model evaluation result is not ideal and the model needs to be adjusted and optimized;

[0092] Select the appropriate model algorithm based on business needs and data characteristics. You can choose a traditional statistical model or a machine learning model, such as decision trees, random forests, neural networks, etc. Use the selected model algorithm to train the data and generate a model. This step requires adjusting the model parameters, optimizing the model's performance, improving the model's prediction accuracy, and evaluating and verifying the trained model, including indicators such as accuracy, recall, and F1 value. Adjust and optimize the model based on the evaluation results to improve the model's performance and stability, deploy the trained model to the production environment, and implement online prediction and monitoring of the model. At the same time, it is necessary to establish a model monitoring mechanism to promptly detect and handle model problems.

[0093] Data modeling is an iterative process that requires continuous adjustment and optimization of model parameters and features to improve model performance and stability. In practical applications, it is also necessary to comprehensively consider issues such as data privacy, security, and compliance to ensure that the application of the model complies with the requirements of relevant laws and regulations;

[0094] Data modeling can realize early warning and assessment of financial risks through the analysis of historical data and real-time data. Through comprehensive analysis of market data, customer data, transaction data, etc., various potential risks can be effectively discovered and assessed, providing decision support and risk control measures for banks. Data modeling can help banks achieve efficient acquisition and processing of diversified and heterogeneous data, ensuring the accuracy and integrity of data. At the same time, through the analysis of risk points and the design of business logic, risks and costs can be reduced and the efficiency of banks can be improved. Data modeling can be applied to audit work. Through data modeling, audit work can ensure that abnormal information can be quickly found, greatly improving audit efficiency. At the same time, the analysis results based on data modeling can improve audit accuracy and reduce the risk of errors caused by non-objective judgment reasons.

[0095] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. These modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be included in the scope of the claims and specification of the present invention.

Claims

1. An online big data intelligent risk control system for the banking industry, characterized by: It includes user information collection module, user classification module, user monitoring module, data processing module and information sending module; The user information collection module is used to collect user information to obtain user information, including user available amount information, user historical consumption times, historical consumption amount, historical consumption location and historical risk control times; The user classification module is used to process the user information to obtain user classification information, and the user classification information includes first-class users, second-class users, third-class users and fourth-class users; The user monitoring module is used to monitor users in real time according to user classification information to obtain real-time user monitoring information; The data processing module is used to process real-time user monitoring information and user classification information to generate user risk control warning information; The information sending module is used to send the user risk control warning information to a preset receiving terminal after the information is generated; The specific process of user classification by the user classification module is as follows: extracting the collected user information, and extracting the user's available amount information, the user's historical consumption times, historical consumption amounts, historical consumption locations and historical risk control times from the user information; The user's available balance information is marked as G1, the user's historical consumption times are marked as G2, the historical consumption amount is marked as G3, and the historical risk control times are marked as G4; Process the historical consumption location information to obtain location evaluation parameters; mark it as G5; The ratio of the historical consumption amount G3 to the user's historical consumption times G2 is calculated to obtain the single consumption evaluation parameter Gg1; When the historical risk control times G4 is less than or equal to the preset value q1, the user's available amount information G1 is greater than the preset value a1, the single consumption evaluation parameter Gg1 is greater than the preset value b1, the user's historical consumption times G2 is greater than the preset value c1, and the location evaluation parameter G5 is greater than the preset value d1, the user is assigned to the second category of users; When the historical risk control times G4 is less than or equal to the preset value q1, the user's available amount information is greater than the preset value a1, the single consumption evaluation parameter Gg1 is between the preset values ​​b1 and b2, the user's historical consumption times G2 is greater than the preset value c1, and the location evaluation parameter G5 is less than or equal to the preset value d1, the user is assigned to the third category; When the historical risk control times G4 is less than or equal to the preset value q1, the user's available amount information is between the preset values ​​a1 and a2, the single consumption evaluation parameter Gg1 is greater than the preset value b1, the user's historical consumption times G2 is greater than the preset value c1, and the location evaluation parameter G5 is less than or equal to the preset value d1, the user is assigned to the third category; When the historical risk control times G4 is greater than the preset value q1, and the user's available amount information G1 is greater than the preset value a1, the user is directly assigned to Category 1 without considering other factors; When the historical risk control times G4 is greater than the preset value q1, and the user's available amount information is between the preset values ​​a1 and a2, that is, without considering other factors, the user is directly assigned to the second category; In addition to the above process, the remaining users are assigned to four categories of users; The specific processing process of processing the historical consumption location information to obtain the location evaluation parameter is as follows: extract the historical consumption location, which is the consumption location information of the x times closest to the current time, import the x times of consumption location information into the preset map software, collect the distance information between the x times of consumption locations, extract the number of distance information with a distance greater than a preset value in the distance information between the x times of consumption locations, and obtain the location evaluation parameter; The specific processing process of the data processing module processing the real-time user monitoring information and user classification information to generate user risk control warning information is as follows: extract the collected user classification information, model the user classification information and import it into a preset risk control model library, determine the classification type of the standard risk control model for user classification, then obtain the real-time user model after modeling the real-time user monitoring information, match the real-time user model with the corresponding classification type, and generate user risk control warning information when a standard risk control model with a similarity greater than a preset value to the real-time user model is matched.

2. The online big data intelligent risk control system for the banking industry according to claim 1, characterized in that: The specific process of user information collection by the user information collection module is as follows: Determine the objectives and scope of data collection and clarify the data that needs to be collected; Select the corresponding data collection method according to the data source and type; Determine the frequency and time of data collection based on business needs and frequency of data changes; Perform data cleaning and preprocessing, including removing duplicate data, filling missing values ​​and processing outliers; Data storage and management: the collected data is stored in a preset data storage device and managed and maintained. The management and maintenance content includes data backup, data archiving, and data security protection.

3. The online big data intelligent risk control system for the banking industry according to claim 1, characterized in that: The specific process of the user monitoring module for real-time monitoring of users according to user classification information to obtain real-time user monitoring information is as follows: Extract user classification information. When the user classification information is a type of user, that is, the user monitoring information is collected every time the user consumes; When the user classification information is a Class II user, the user monitoring information is collected after the user consumes at least e1 times; When the user classification information is the third category of users, the user monitoring information is collected after the user's consumption times reaches e2; When the user classification information is four types of users, the user monitoring information is collected every preset time period.

4. The online big data intelligent risk control system for the banking industry according to claim 1, characterized in that: The risk control model library has preset one type of risk control model, two type of risk control model, three type of risk control model and four type of risk control model; One type of risk control model corresponds to one type of user, the second type of risk control model corresponds to the second type of user, the third type of risk control model corresponds to the third type of user, and the fourth type of risk control model corresponds to the fourth type of user.

5. The online big data intelligent risk control system for the banking industry according to claim 4, characterized in that: The specific process of obtaining the real-time user model after modeling the real-time user monitoring information is as follows: extracting the collected real-time user monitoring information, the real-time user monitoring information includes consumption time, consumption location, consumption amount, and consumption frequency, and then extracting features from the real-time user monitoring information, and extracting indicators reflecting the characteristics of user consumption behavior by analyzing the user's consumption data, including consumption frequency, average consumption amount, and consumption location distribution, and then using the extracted features to build a variety of risk control models, including a real-time logistic regression model, a real-time decision tree model, and a real-time random forest model; In the process of matching the real-time user model with the corresponding classification type, the established real-time logistic regression model, real-time decision tree model and real-time random forest model are synchronized and matched with the models in the preset risk control model library.

6. The online big data intelligent risk control system for the banking industry according to claim 5, characterized in that: After the real-time logistic regression model, the real-time decision tree model and the real-time random forest model are established, it is also necessary to perform model evaluation and optimization on the real-time logistic regression model, the real-time decision tree model and the real-time random forest model.

7. The online big data intelligent risk control system for the banking industry according to claim 6, characterized in that: The determination process of model evaluation and optimization for the real-time logistic regression model, the real-time decision tree model and the real-time random forest model is as follows: the accuracy, recall rate and F1 value of the real-time logistic regression model, the real-time decision tree model and the real-time random forest model are collected. When the accuracy, recall rate and F1 value do not meet the preset conditions, it means that the model evaluation result is not ideal and the model needs to be adjusted and optimized.