Credit and loan scene risk identification method and equipment based on big data model, and medium

Abstract

The invention discloses a credit scene risk identification method and device based on a big data model, and a medium, and relates to the technical field of risk identification, and the method comprises the steps: obtaining a safety credit feature based on the historical debit and credit data of a safety credit website; acquiring risk data features by using a security credit analysis method; establishing a credit analysis model; performing risk identification based on a credit analysis model; the method is used for solving the problems that in an existing credit scene risk identification method, when the number of users and the number of transactions of a website are large, due to the fact that the updating frequency of user data and website data in the website is high, the extraction speed of the user data is low, and the risk identification speed is low; therefore, the problem that the loan risk of the website cannot be effectively and accurately identified in real time based on the loan data of the user is solved.

Description

Technical Field

[0001] The present invention relates to the field of risk identification technology, and specifically to a credit scenario risk identification method, device, and medium based on a big data model. Background Art

[0002] Credit scenarios refer to various specific scenarios in which financial services are integrated into social production and social life, including both online and offline methods; online scenarios include online loan platforms and mobile payment applications, while offline scenarios include traditional bank branches and merchant cooperation; risk identification in credit scenarios usually includes identification of credit risk, market risk, operational risk, liquidity risk, legal risk and related enterprise risk.

[0003] Existing risk identification methods for credit scenarios usually establish a model, use the user's loan data, economic status data and transaction behavior data to train the model, and perform segmented training through the existing network model to improve the model's credit risk prediction ability for user data with different data distributions. Although this improved method can predict credit risks for users with different data states, if a model is established for each user and all data of each user is analyzed based on segmented training after data acquisition, when the number of users and transaction volume of the website is large, when risk identification is performed for all users, there will be problems such as slow user data extraction speed and slow risk identification speed through data training, resulting in the problem of being unable to effectively and accurately identify the website's loan risk in real time based on the user's loan data. For example, in the patent application with publication number CN119941382A, a credit risk identification model is disclosed. Training method, credit risk identification method and device. This solution is to train the initial credit risk identification model by using relevant user data, generate training sample data with different data distributions to train the classic network sub-model, thereby improving the credit risk prediction ability of the classic network sub-model for sample data with different data distributions. Other improvements in risk identification for credit scenarios are usually based on improvements in the accuracy of risk identification based on community structure, which still cannot solve the problem that when the number of users and transaction volume of the website is large, the user data and website data in the website are updated frequently, resulting in slow extraction speed of user data and slow risk identification speed, resulting in the inability to effectively and accurately identify the website's lending risk based on the user's lending data in real time. In view of this, it is necessary to improve the existing risk identification methods for credit scenarios. Summary of the Invention

[0004] The present invention aims to solve one of the technical problems in the prior art to at least a certain extent, by proposing a credit scenario risk identification method, device and medium based on a big data model, so as to solve the problem in the existing credit scenario risk identification method that when the number of users and transaction volume of the website is large, when risk identification of all users or analysis of the community structure is performed, the user data and website data in the website are updated frequently, resulting in a slow extraction speed of user data and a slow identification speed of risk, thereby making it impossible to effectively and accurately identify the website's lending risk in real time based on the user's lending data.

[0005] To achieve the above objectives, in a first aspect, the present application provides a credit scenario risk identification method based on a big data model, comprising the following steps: The online platform where the credit scenario occurs is recorded as the credit platform; based on the loan type of the credit platform, safe credit websites are obtained using big data, and the historical loan data of the safe credit websites is analyzed. Based on the analysis results, safe credit characteristics are obtained, where the safe credit characteristics include safe lending areas and safe return areas; Analyze the historical lending data of the credit platform using the safe credit analysis method based on the safe credit characteristics, and obtain the risk data characteristics of the credit platform based on the analysis results; A big data model is established using the historical lending data of a safe credit website as a data source, and recorded as a credit analysis model; the credit analysis model is used to conduct real-time analysis of the historical data of the safe credit website, and the safe credit characteristics are updated in real time; the risk data characteristics of the lending platform are obtained in real time based on the latest safe credit characteristics, and recorded as real-time risk characteristics; when a credit platform has real-time risk characteristics, the credit platform is recorded as a risk platform and the real-time risk characteristics are used as the risk data of the credit platform.

[0006] Furthermore, based on the loan types of the credit platform, we use big data to obtain safe credit websites and analyze the historical loan data of the safe credit websites. Based on the analysis results, we obtain safe credit features including: Based on big data, multiple credit websites identified as risk-free are obtained and recorded as safe credit websites. For any safe credit website, the historical lending data of the safe credit website is recorded as safe lending data, and the safe credit data is analyzed using a lending analysis method. Based on the analysis results, the lending feature area and the return feature area of ​​the safe credit website are obtained. The union of the lending feature areas of all safe credit websites is recorded as the safe lending area, and the union of the return feature areas of all safe credit websites is recorded as the safe return area.

[0007] Furthermore, the loan analysis method includes: obtaining all loan records in the secure credit data and recording them as secure loan records; for any secure loan record, recording the amount of each loan in the secure loan record as a secure loan amount, and recording the time corresponding to when each secure loan amount is lent and the time corresponding to when the loan is repaid as a secure loan time and a secure return time, wherein the secure loan time and the secure return time are both times consisting only of hours, minutes, and seconds; Get the safe loan amount and the corresponding safe loan time and safe return time in all safe loan records; Sort all safe loan amounts from small to large and record them as safe loan amount AJ1 to safe loan amount AJ n .

[0008] Furthermore, the loan analysis method also includes: Create a table with R rows and 4 columns and record it as a loan analysis table. In the first row of the loan analysis table, except for the leftmost cell, fill in the loan time, return time, and number of times from left to right. In the leftmost column of the loan analysis table, except for the top cell, fill in the safe loan time AJ1 to safe loan time AJ from top to bottom. n ; In the loan analysis table: For any safe loan amount AJ c , the safe loan amount AJ c All corresponding safe lending times are recorded in the behavioral safe lending time AJ c And the cell α is listed as the loan time, and the safe loan amount AJ c All corresponding safe return times are recorded in the behavioral safe loan time AJ c And the cell β is listed as the return time, and the sum of the time recorded in cell α and cell β is recorded in the behavioral security loan time AJ c And the cells are listed as times, where the same safe loan amount can correspond to multiple safe loan times and multiple safe return times.

[0009] Furthermore, the loan analysis method also includes: Establish a plane rectangular coordinate system, denoted as the credit extraction coordinate system, where the unit of the X-axis of the credit extraction coordinate system is yuan and the unit of the Y-axis is h; for row δ between rows 2 and n+1 in the credit analysis table, denote the data in the cells in row δ from left to right as A1, A2, A3, and A4, where δ is a positive integer less than or equal to n+1 and greater than or equal to 2; With A1 as the horizontal coordinate, all the times in A2 are marked in the credit withdrawal coordinate system as the vertical coordinates, and all the marks are recorded as loan marks; with A1 as the horizontal coordinate, all the times in A2 are marked in the credit withdrawal coordinate system as the vertical coordinates, and all the marks are recorded as return marks; when A4 in the δth row is less than A4 in the δ-1th row and less than A4 in the δ+th row, the δth row is recorded as a valley row; when A4 in the δth row is greater than A4 in the δ-1th row and greater than A4 in the δ+th row, the δth row is recorded as a peak row; All rows between the 2nd row and the n+1th row are processed based on the processing method for the δth row, and after processing all rows, only the lending punctuation marks and return punctuation marks corresponding to all valley rows and peak rows are retained in the credit extraction coordinate system. Among them, when the 2nd row and the n+1th row are not recorded as valley rows or peak rows, the lending punctuation marks and return punctuation marks corresponding to the 2nd row and the n+1th row are still retained in the credit extraction coordinate system.

[0010] Furthermore, the loan analysis method also includes: For any lending point with a horizontal coordinate X1, the lending point with the horizontal coordinate X1 and the largest and smallest vertical coordinates is recorded as the lending high point and the lending low point; the return point with the horizontal coordinate X1 and the largest and smallest vertical coordinates is recorded as the return high point and the return low point; The curve obtained by fitting all the high points of lending is recorded as the high-borrowing curve, the curve obtained by fitting all the low points of lending is recorded as the low-borrowing curve, the curve obtained by fitting all the high points of return is recorded as the high-returning curve, and the curve obtained by fitting all the low points of return is recorded as the low-returning curve. Among them, the leftmost point and the rightmost point of the high-borrowing curve, the low-borrowing curve, the high-returning curve and the low-returning curve are all the lending punctuation points or the returning punctuation points. The area between the loaned high song and the loaned low song is recorded as the loan feature area; the area between the returned high song and the returned low song is recorded as the return feature area.

[0011] Furthermore, based on the safe credit characteristics, the safe credit analysis method is used to analyze the historical lending data of the credit platform, and the risk data characteristics of the credit platform are obtained based on the analysis results, including: Obtain all loan records from the lending platform's historical lending data and record them as pending verification records. For any pending verification record, use the amount in the pending verification record as the horizontal axis and the loan time and repayment time as the vertical axis to mark them in the credit extraction coordinate system, and record them as the pending verification loan point and pending verification repayment point, respectively. When the borrowing point to be verified is in a safe lending area and the return point to be verified is in a safe return area, the record to be verified will be recorded as a risk-free record.

[0012] Furthermore, step S2 further includes: When the loan point to be verified is in the safe lending area, and the return point to be verified is not in the safe return area, the record to be verified will be recorded as a return risk record; When the loan point to be verified is not in the safe lending area, and the return point to be verified is in the safe return area, the record to be verified will be recorded as a lending risk record; When the borrowing point to be verified is not in the safe lending area, and the return point to be verified is not in the safe return area, the record to be verified will be recorded as a double risk record; All return risk records, lending risk records, and double risk records are recorded as risk data features.

[0013] In a second aspect, the present application provides an electronic device comprising a processor and a memory, wherein the memory stores computer-readable instructions. When the computer-readable instructions are executed by the processor, the steps in the above method are performed.

[0014] In a third aspect, the present application provides a storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps in the above method are performed.

[0015] Beneficial effects of the present invention: The present application first uses big data to obtain safe credit websites based on the loan types of the credit platform, analyzes the historical loan data of the safe credit websites, and obtains safe credit features based on the analysis results; then, based on the safe credit features, uses a safe credit analysis method to analyze the historical loan data of the credit platform, and obtains the risk data features of the credit platform based on the analysis results. The advantage of this is that, by obtaining safe credit features based on safe credit websites and obtaining risk data features through safe credit features, after obtaining features corresponding to the transaction status of risk-free credit websites, features of the transaction status on the credit platform that violate the safe credit features can be obtained based on the historical loan data of the credit platform. This allows, after subsequently establishing a credit analysis model, to quickly obtain the latest risk data features of the credit platform based on the real-time updated data of the credit platform, thereby performing real-time and efficient risk identification of the credit platform. This application also establishes a big data model with the historical lending data of the safe credit website as the data source, and records it as a credit analysis model; uses the credit analysis model to perform real-time analysis on the historical data of the safe credit website, and updates the safe credit characteristics in real time; obtains the risk data characteristics of the lending platform in real time based on the latest safe credit characteristics, and records them as real-time risk characteristics; when the credit platform has real-time risk characteristics, the credit platform is recorded as a risk platform and the real-time risk characteristics are used as the risk data of the credit platform. The advantage of this is that by establishing a credit analysis model, the real-time data of the safe credit website and the real-time data of the credit platform can be automatically captured respectively, thereby updating the safe credit characteristics and the risk data characteristics of the credit platform in real time. Because after the data is automatically captured, only the transaction time and transaction amount of the loan are analyzed, the speed of data analysis can be effectively improved, so that when the number of users and transaction volume of the website is large, the speed of extracting and analyzing user data and the speed of identifying risks of the credit platform can be improved, so as to realize real-time, efficient and accurate identification of the website's lending risks based on user lending data. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 is a flow chart of the steps of the method of the present invention; Figure 2 A schematic diagram of marking the lending punctuation marks and the returning punctuation marks of the present invention; Figure 3 A schematic diagram of obtaining a lending feature area and a returning feature area according to the present invention; Figure 4 Schematic diagram of the structure of the electronic device of the present invention. DETAILED DESCRIPTION

[0017] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0018] Example 1, please refer to Figure 1 As shown, this application provides a credit scenario risk identification method based on a big data model, comprising the following steps: Step S1: Record the network platform where the credit scenario occurs as a credit platform; use big data to obtain safe credit websites based on the loan types of the credit platform, analyze the historical loan data of the safe credit websites, and obtain safe credit characteristics based on the analysis results, wherein the safe credit characteristics include a safe lending area and a safe return area; During the specific implementation process, loan types may include bank loans, guaranteed loans, and mortgage loans; safe credit websites can be obtained based on the loan types corresponding to the credit scenarios for risk identification to ensure that the safe credit features obtained by the safe credit websites can effectively identify the risks of the credit platform.

[0019] Step S1 includes: Step S101, based on big data, obtaining multiple credit websites identified as risk-free and recording them as safe credit websites; for any safe credit website: recording the historical lending data of the safe credit website as safe lending data, analyzing the safe credit data using a lending analysis method, and obtaining a lending feature area and a return feature area of ​​the safe credit website based on the analysis results; The loan analysis method includes: step S1011, obtaining all loan records in the secure credit data and recording them as secure loan records; for any secure loan record, recording the amount of each loan in the secure loan record as the secure loan amount, and recording the time corresponding to when each secure loan amount is lent and the time corresponding to when the loan is repaid as the secure loan time and the secure return time, wherein the secure loan time and the secure return time are both times consisting only of hours, minutes, and seconds; Get the safe loan amount and the corresponding safe loan time and safe return time in all safe loan records; During the specific implementation process, for example, during a data analysis, a safe loan record is obtained in which the loan amount is 1,000 yuan, and the loan time and repayment time are 14:30:00 on October 20, 2020 and 12:30:00 on November 20, 2020, respectively. The safe loan time and safe return time can be recorded as 14:30:00 and 12:30:00, respectively. By obtaining the safe loan time and safe return time of each safe loan record, the time distribution of the loan time and return time corresponding to different loan amounts on the safe credit website can be obtained, so as to identify the risks of the credit platform based on the loan amount and the corresponding loan time and return time in subsequent analysis; Step S1012: sort all the safe loan amounts from small to large and record them as safe loan amount AJ1 to safe loan amount AJ n .

[0020] In the specific implementation process, for example, in a data analysis, the three safe credit records obtained have safe loan amounts of 1,000 yuan, 2,000 yuan, 3,000 yuan, 800 yuan, 1,000 yuan, 5,000 yuan and 2,000 yuan, 3,000 yuan, and 10,000 yuan respectively. Then, through data analysis, it can be obtained that the safe loan amount AJ1 to the safe loan amount AJ nThey are 800 yuan, 1000 yuan, 2000 yuan, 3000 yuan, 5000 yuan and 10000 yuan respectively, and the value of n is 6; Step S1013: Create a table with R rows and 4 columns and record it as a loan analysis table. In the first row of the loan analysis table, except for the leftmost cell, fill in the loan time, return time, and number of times from left to right. In the leftmost column of the loan analysis table, except for the top cell, fill in the safe loan time AJ1 to safe loan time AJ from top to bottom. n ; Step S1014, in the loan analysis table: for any safe loan amount AJ c , the safe loan amount AJ c All corresponding safe lending times are recorded in the behavioral safe lending time AJ c And the cell α is listed as the loan time, and the safe loan amount AJ c All corresponding safe return times are recorded in the behavioral safe loan time AJ c And the cell β is listed as the return time, and the sum of the time recorded in cell α and cell β is recorded in the behavioral security loan time AJ c And the cells are listed as times, where the same safe loan amount can correspond to multiple safe loan times and multiple safe return times; In a specific real-time process, for example, during a data analysis, the obtained loan analysis table is shown in Table 1: Loan Analysis Table; Table 1: Loan Analysis Table Lend Time Return time frequency 800 14:30:00 12:30:00 2 1000 11:00:00；12:00:00 12:00:00；10:30:00 4 2000 17:30:00；18:00:00 14:30:00；14:30:00 4 3000 12:30:00 9:30:00 2 5000 5:30:00；6:00:00 7:30:00；7:00:00 4 10000 1:00:00 23:30:00 2 Step S1015: Establish a plane rectangular coordinate system, denoted as the credit extraction coordinate system, where the unit of the X-axis of the credit extraction coordinate system is yuan and the unit of the Y-axis is h. For row δ between rows 2 and n+1 in the credit analysis table, denote the data in the cells in row δ from left to right as A1, A2, A3, and A4, where δ is a positive integer less than or equal to n+1 and greater than or equal to 2. In a specific implementation, for example, a loan analysis table obtained during a data analysis is shown in Table 1: Loan Analysis Table. For row 4 in Table 1, A1, A2, A3, and A4 are: 2000, 17:30:00; 18:00:00, 14:30:00; 14:30:00, and 4, respectively; Step S1016: Mark the credit withdrawal coordinate system with A1 as the horizontal coordinate and all the times in A2 as the vertical coordinates, and record all the marks as loan marks; mark the credit withdrawal coordinate system with A1 as the horizontal coordinate and all the times in A2 as the vertical coordinates, and record all the marks as return marks; when A4 in the δth row is less than A4 in the δ-1th row and less than A4 in the δ+th row, record the δth row as a valley row; when A4 in the δth row is greater than A4 in the δ-1th row and greater than A4 in the δ+th row, record the δth row as a peak row; Step S1017: Process all rows between row 2 and row (n+1) based on the same processing method as for row δ. After processing all rows, only the lending and returning punctuation marks corresponding to all valley rows and peak rows are retained in the credit withdrawal coordinate system. If row 2 and row (n+1) are not recorded as valley rows or peak rows, the lending and returning punctuation marks corresponding to row 2 and row (n+1) are still retained in the credit withdrawal coordinate system. In the specific implementation process, for example, in a data analysis, based on the analysis of Table 1, it can be obtained that the 5th row in Table 1 is the valley row and the 6th row is the peak row. Then, by retaining the lending punctuation and return punctuation corresponding to the 2nd row, the 5th row, the 6th row and the 7th row, the credit withdrawal coordinate system is obtained as follows: Figure 2 As shown, △ is the borrowing punctuation mark, ○ is the returning punctuation mark; Figure 2 The borrowing high point, borrowing low point, return high point and return low point are extracted, and the borrowing high song, borrowing low song, return high song and return low song are obtained as follows: Figure 3 As shown by the curves JG, JD, GG, and GD in FIG, the area between the curves JG and JD can be recorded as the lending feature area, and the area between the curves GG and GD can be recorded as the returning feature area; Step S1019: For any lending point with a horizontal coordinate X1, the lending point with the horizontal coordinate X1 and the largest and smallest vertical coordinates is recorded as the lending high point and the lending low point; the return point with the horizontal coordinate X1 and the largest and smallest vertical coordinates is recorded as the return high point and the return low point; The curve obtained by fitting all the high points of lending is recorded as the high-borrowing curve, the curve obtained by fitting all the low points of lending is recorded as the low-borrowing curve, the curve obtained by fitting all the high points of return is recorded as the high-returning curve, and the curve obtained by fitting all the low points of return is recorded as the low-returning curve. Among them, the leftmost point and the rightmost point of the high-borrowing curve, the low-borrowing curve, the high-returning curve and the low-returning curve are all the lending punctuation points or the returning punctuation points. The area between the loaned high song and the loaned low song is recorded as the loan feature area; the area between the returned high song and the returned low song is recorded as the returned feature area; During the specific implementation process, by obtaining the lending feature area and the repayment feature area, the characteristic intervals of the lending time and repayment time corresponding to each loan amount under the condition of safe lending can be obtained, so as to provide data support for the acquisition of risk data characteristics of the credit platform.

[0021] In step S102, the union of the lending feature areas of all the safe credit websites is recorded as a safe lending area, and the union of the return feature areas of all the safe credit websites is recorded as a safe return area.

[0022] Step S2: Analyze the historical lending data of the credit platform using a safe credit analysis method based on the safe credit characteristics, and obtain the risk data characteristics of the credit platform based on the analysis results; Step S2 includes: Step S201, obtaining all loan records in the historical loan data of the loan platform and recording them as pending verification records; for any pending verification record, punctuating the credit extraction coordinate system with the amount in the pending verification record as the horizontal axis and the loan time and return time as the vertical axis, and recording them as pending verification loan points and pending verification return points respectively; Step S202: When the loan point to be verified is in a safe lending area and the return point to be verified is in a safe return area, the record to be verified is recorded as a risk-free record; Step S203: When the loan point to be verified is in the safe lending area and the return point to be verified is not in the safe return area, the record to be verified is recorded as a return risk record; Step S204: When the loan point to be verified is not in the safe lending area, and the return point to be verified is in the safe return area, the record to be verified is recorded as a lending risk record; Step S205: When the loan point to be verified is not in the safe lending area, and the return point to be verified is not in the safe return area, the record to be verified is recorded as a double risk record; Step S206: record all return risk records, lending risk records, and dual risk records as risk data features.

[0023] Step S3: Establish a big data model using the historical lending data of the safe credit website as a data source, and record it as a credit analysis model; use the credit analysis model to perform real-time analysis on the historical data of the safe credit website, and update the safe credit characteristics in real time; obtain the risk data characteristics of the lending platform in real time based on the latest safe credit characteristics, and record them as real-time risk characteristics; when a credit platform has real-time risk characteristics, record the credit platform as a risk platform and use the real-time risk characteristics as the risk data of the credit platform.

[0024] Example 2, please refer to Figure 4 As shown, Figure 4A schematic diagram of the structure of an electronic device is provided. The electronic device may include: a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other via the communication bus. The memory stores computer-readable instructions, and the processor can call the instructions in the memory. When the computer-readable instructions are executed by the processor, the steps of a credit scenario risk identification method based on a big data model are executed to implement the following functions: first, based on the loan type of the credit platform, safe credit websites are obtained using big data, and historical loan data of the safe credit websites is analyzed, and safe credit characteristics are obtained based on the analysis results; then, based on the safe credit characteristics, the historical loan data of the credit platform is analyzed using a safe credit analysis method, and risk data characteristics of the credit platform are obtained based on the analysis results; finally, a big data model is established using the historical loan data of the safe credit website as a data source, and recorded as a credit analysis model; the credit analysis model is used to analyze the historical data of the safe credit website in real time, and the safe credit characteristics are updated in real time; risk data characteristics of the loan platform are obtained in real time based on the latest safe credit characteristics, and recorded as real-time risk characteristics; when a credit platform has real-time risk characteristics, the credit platform is recorded as a risk platform, and the real-time risk characteristics are used as the risk data of the credit platform.

[0025] In addition, the logical instructions in the above-mentioned memory can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on this understanding, the technical solution of the present application, or the part that contributes to the existing technology, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, and other media that can store program code.

[0026] Example 3. The present application also provides a computer program product, which includes a computer program stored on a computer-readable storage medium, and the computer program includes program instructions. When the program instructions are executed by a computer, the computer can execute the credit scenario risk identification method based on the big data model provided by the above methods, the method including: first, using big data to obtain a safe credit website based on the loan type of the credit platform, and analyzing the historical loan data of the safe credit website, and obtaining safe credit characteristics based on the analysis results; then, using a safe credit analysis method to analyze the historical loan data of the credit platform based on the safe credit characteristics, and obtaining the risk data characteristics of the credit platform based on the analysis results; finally, establishing a big data model with the historical loan data of the safe credit website as the data source, and recording it as a credit analysis model; using the credit analysis model to perform real-time analysis on the historical data of the safe credit website, and updating the safe credit characteristics in real time; obtaining the risk data characteristics of the lending platform in real time based on the latest safe credit characteristics, and recording them as real-time risk characteristics; when the credit platform has real-time risk characteristics, the credit platform is recorded as a risk platform and the real-time risk characteristics are used as the risk data of the credit platform.

[0027] Example 4. The present application also provides a computer-readable storage medium. The present application provides a storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps in the credit scenario risk identification method based on the big data model are executed to achieve the following functions: first, based on the lending type of the credit platform, big data is used to obtain a safe credit website, and the historical lending data of the safe credit website is analyzed, and safe credit characteristics are obtained based on the analysis results; then, based on the safe credit characteristics, the historical lending data of the credit platform is analyzed using a safe credit analysis method, and the risk data characteristics of the credit platform are obtained based on the analysis results; finally, a big data model is established with the historical lending data of the safe credit website as the data source, and recorded as a credit analysis model; the credit analysis model is used to perform real-time analysis on the historical data of the safe credit website, and the safe credit characteristics are updated in real time; based on the latest safe credit characteristics, the risk data characteristics of the lending platform are obtained in real time, and recorded as real-time risk characteristics; when the credit platform has real-time risk characteristics, the credit platform is recorded as a risk platform and the real-time risk characteristics are used as the risk data of the credit platform.

[0028] Through the description of the above embodiments, the embodiments of the present invention can be provided as methods, systems, or computer program products. Based on this understanding, the essence of the above technical solutions or the portion that contributes to the prior art can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, or an optical disk, and includes a number of instructions for enabling a computer device (such as a personal computer, server, or network device) to execute the methods described in various embodiments or certain portions of the embodiments.

[0029] In the embodiments provided in this application, it should be understood that the disclosed system or method can be implemented in other ways. The embodiments described above are merely illustrative. For example, the division of modules or units is only a logical function division. There may be other division methods in actual implementation. For example, multiple modules or units can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interfaces, and the indirect coupling or communication connection of systems, modules and units can be electrical, mechanical or other forms.

[0030] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application 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 make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A credit scenario risk identification method based on a big data model, characterized by: The steps include: The online platform where the credit scenario occurs is recorded as the credit platform; based on the loan type of the credit platform, safe credit websites are obtained using big data, and the historical loan data of the safe credit websites is analyzed. Based on the analysis results, safe credit characteristics are obtained, where the safe credit characteristics include safe lending areas and safe return areas; Analyze the historical lending data of the credit platform using the safe credit analysis method based on the safe credit characteristics, and obtain the risk data characteristics of the credit platform based on the analysis results; A big data model is established using the historical lending data of a safe credit website as a data source, and recorded as a credit analysis model; the credit analysis model is used to conduct real-time analysis of the historical data of the safe credit website, and the safe credit characteristics are updated in real time; the risk data characteristics of the lending platform are obtained in real time based on the latest safe credit characteristics, and recorded as real-time risk characteristics; when a credit platform has real-time risk characteristics, the credit platform is recorded as a risk platform and the real-time risk characteristics are used as the risk data of the credit platform.

2. The credit scenario risk identification method based on big data model according to claim 1 is characterized in that: Based on the loan types of the credit platform, we use big data to obtain safe credit websites and analyze the historical loan data of safe credit websites. Based on the analysis results, we obtain safe credit features including: Based on big data, multiple credit websites identified as risk-free are obtained and recorded as safe credit websites. For any safe credit website, the historical lending data of the safe credit website is recorded as safe lending data, and the safe credit data is analyzed using a lending analysis method. Based on the analysis results, the lending feature area and the return feature area of ​​the safe credit website are obtained. The union of the lending feature areas of all safe credit websites is recorded as the safe lending area, and the union of the return feature areas of all safe credit websites is recorded as the safe return area.

3. The credit scenario risk identification method based on big data model according to claim 2 is characterized in that: The loan analysis method includes: obtaining all loan records in the secure credit data and recording them as secure loan records; for any secure loan record, recording the amount of each loan in the secure loan record as the secure loan amount, and recording the time corresponding to when each secure loan amount is borrowed and the time corresponding to when the loan is repaid as the secure loan time and the secure return time, wherein the secure loan time and the secure return time are both times consisting only of hours, minutes, and seconds; Get the safe loan amount and the corresponding safe loan time and safe return time in all safe loan records; Sort all safe loan amounts from small to large and record them as safe loan amount AJ1 to safe loan amount AJ n .

4. The credit scenario risk identification method based on big data model according to claim 3 is characterized in that: Borrowing and lending analysis methods also include: Create a table with R rows and 4 columns and record it as a loan analysis table. In the first row of the loan analysis table, except for the leftmost cell, fill in the loan time, return time, and number of times from left to right. In the leftmost column of the loan analysis table, except for the top cell, fill in the safe loan time AJ1 to safe loan time AJ from top to bottom. n ; In the loan analysis table: For any safe loan amount AJ c , the safe loan amount AJ c All corresponding safe lending times are recorded in the behavioral safe lending time AJ c And the cell α is listed as the loan time, and the safe loan amount AJ c All corresponding safe return times are recorded in the behavioral safe loan time AJ c And the cell β is listed as the return time, and the sum of the time recorded in cell α and cell β is recorded in the behavioral security loan time AJ c And the cells are listed as times, where the same safe loan amount can correspond to multiple safe loan times and multiple safe return times.

5. The credit scenario risk identification method based on big data model according to claim 4 is characterized in that: Borrowing and lending analysis methods also include: Establish a plane rectangular coordinate system, denoted as the credit extraction coordinate system, where the unit of the X-axis of the credit extraction coordinate system is yuan and the unit of the Y-axis is h; for row δ between rows 2 and n+1 in the credit analysis table, denote the data in the cells in row δ from left to right as A1, A2, A3, and A4, where δ is a positive integer less than or equal to n+1 and greater than or equal to 2; With A1 as the horizontal coordinate, all the times in A2 are marked in the credit withdrawal coordinate system as the vertical coordinates, and all the marks are recorded as loan marks; with A1 as the horizontal coordinate, all the times in A2 are marked in the credit withdrawal coordinate system as the vertical coordinates, and all the marks are recorded as return marks; when A4 in the δth row is less than A4 in the δ-1th row and less than A4 in the δ+th row, the δth row is recorded as a valley row; when A4 in the δth row is greater than A4 in the δ-1th row and greater than A4 in the δ+th row, the δth row is recorded as a peak row; All rows between the 2nd row and the n+1th row are processed based on the processing method for the δth row, and after processing all rows, only the lending punctuation marks and return punctuation marks corresponding to all valley rows and peak rows are retained in the credit extraction coordinate system. Among them, when the 2nd row and the n+1th row are not recorded as valley rows or peak rows, the lending punctuation marks and return punctuation marks corresponding to the 2nd row and the n+1th row are still retained in the credit extraction coordinate system.

6. The credit scenario risk identification method based on big data model according to claim 5 is characterized in that: Borrowing and lending analysis methods also include: For any lending point with a horizontal coordinate X1, the lending point with the horizontal coordinate X1 and the largest and smallest vertical coordinates is recorded as the lending high point and the lending low point; the return point with the horizontal coordinate X1 and the largest and smallest vertical coordinates is recorded as the return high point and the return low point; The curve obtained by fitting all the high points of lending is recorded as the high-borrowing curve, the curve obtained by fitting all the low points of lending is recorded as the low-borrowing curve, the curve obtained by fitting all the high points of return is recorded as the high-returning curve, and the curve obtained by fitting all the low points of return is recorded as the low-returning curve. Among them, the leftmost point and the rightmost point of the high-borrowing curve, the low-borrowing curve, the high-returning curve and the low-returning curve are all the lending punctuation points or the returning punctuation points. The area between the loaned high song and the loaned low song is recorded as the loan feature area; the area between the returned high song and the returned low song is recorded as the return feature area.

7. The credit scenario risk identification method based on big data model according to claim 6 is characterized in that: Based on the safety credit characteristics, the safety credit analysis method is used to analyze the historical lending data of the credit platform, and the risk data characteristics of the credit platform are obtained based on the analysis results, including: Obtain all loan records from the lending platform's historical lending data and record them as pending verification records. For any pending verification record, use the amount in the pending verification record as the horizontal axis and the loan time and repayment time as the vertical axis to mark them in the credit extraction coordinate system, and record them as the pending verification loan point and pending verification repayment point, respectively. When the borrowing point to be verified is in a safe lending area and the return point to be verified is in a safe return area, the record to be verified will be recorded as a risk-free record.

8. The credit scenario risk identification method based on big data model according to claim 7 is characterized in that: Step S2 further includes: When the loan point to be verified is in the safe lending area, and the return point to be verified is not in the safe return area, the record to be verified will be recorded as a return risk record; When the loan point to be verified is not in the safe lending area, and the return point to be verified is in the safe return area, the record to be verified will be recorded as a lending risk record; When the borrowing point to be verified is not in the safe lending area, and the return point to be verified is not in the safe return area, the record to be verified will be recorded as a double risk record; All return risk records, lending risk records, and double risk records are recorded as risk data features.

9. An electronic device, characterized in that: The method comprises a processor and a memory, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the steps in the method according to any one of claims 1 to 8 are executed.

10. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 8 are executed.

Images