CA3141554A1

CA3141554A1 - User credit risk assessment method, device, computer equipment and storage medium

Info

Publication number: CA3141554A1
Application number: CA3141554A
Authority: CA
Inventors: Hang QIAN
Original assignee: 10353744 Canada Ltd
Current assignee: 10353744 Canada Ltd
Priority date: 2020-12-09
Filing date: 2021-12-09
Publication date: 2022-06-09
Also published as: CA3177674A1; CN112613971A

Abstract

The present invention makes public a method of and a device for assessing user credit risk, a computer equipment and a storage medium. The method comprises: obtaining a first user data of a target user on a first target platform and a second user data of the target user on a second target platform, employing a preset algorithm to analyze the first user data, and obtaining a target variable to which the target user corresponds, determining an initial admission condition of credit according to the target variable and a preset threshold, readjusting the initial admission condition according to the second user data, and obtaining a target admission condition, and determining a credit line of each target user that conforms to the target admission condition according to the first user data and the second user data.

Description

USER CREDIT RISK ASSESSMENT METHOD, DEVICE, COMPUTER EQUIPMENT
AND STORAGE MEDIUM
BACKGROUND OF THE INVENTION
Technical Field [0001] The present invention relates to the field of data processing technology, and more particularly to a method of and a device for assessing user credit risk, a computer equipment and a storage medium.
Description of Related Art

[0002] It is an important national policy of our country in the current phase to issue credit loans to small and micro enterprises, and this is also an important undertaking to help the small and micro enterprises and to make prosperous the numerous small and micro enterprises of our country. There are indeed numerous small and micro enterprises, according to data made by the State Bureau of Statistics, there are currently about seventy million active small and micro enterprises in China today. A large number of small and micro enterprises requires the support of credit loans, and the credit loan market of small and micro enterprises is extensive. There are many modes to supply credit loans to small and micro enterprises, such as the traditional credit mode of mortgage and pledge by clients visits, the business mode of offline investigation of clients by customer managers on a one-by-one basis, the credit mode without mortgage with online and offline services combined, the purely online business mode carried out by means of tax data, and the business mode to issue credit loans to downstream clients according to information possessed by a core enterprise of the supply chain.

[0003] All the aforementioned credit modes issue credit loans mainly through two methods, one is to acquire information of sufficient mortgages and pledges of clients, once there are sufficient mortgages and pledges, it will be unnecessary for the credit institution to fully learn the operational circumstances of clients, and it will be also unnecessary to worry about the loans to become bad loans; the other method is to fully learn the operational information of clients, Date recue / Date received 2021-12-09 frilly assess the operational circumstances of clients, and judge the levels of credit risks of enterprises, before credit loans are issued to the enterprises.

[0004] However, both of the aforementioned two credit methods are more or less defective. As regards the first method, since it is impossible to acquire the complete information, operational information in particular, of the clients, the bad loan rate is high in the client group credits of those core enterprises whose control over downstream clients is not quite well. In the second method, although the data collected offline can be complete, the time taken is long, the credit cost is high, and it is further required for some credit institutions to make use of online tax data, while the mode of processing tax data is usually coarse, and the effect thereof is meager.

[0005] In short, there is an urgent need to propose a novel method of assessing user credit risk, so as to overcome the aforementioned problems.
SUMMARY OF THE INVENTION

[0006] In order to address the technical problems pending in the state of the art, embodiments of the present invention provide a method of and a device for assessing user credit risk, a computer equipment and a storage medium, so as to overcome such problems prevailing in prior-art technology as high bad loan rate of credit caused by incomplete client information, long time duration in collecting client information, and high credit cost, etc.

[0007] To solve one or more of the aforementioned technical problem(s), the present invention employs the following technical solutions.

[0008] According to the first aspect, there is provided a method of assessing user credit risk, and the method comprises the following steps:

[0009] obtaining a first user data of a target user on a first target platform and a second user data of the target user on a second target platform;

Date recue / Date received 2021-12-09

[0010] employing a preset algorithm to analyze the first user data, and obtaining a target variable to which the target user corresponds, wherein the target variable includes a variable describing credit risk of the target user;

[0011] determining an initial admission condition of credit according to the target variable and a preset threshold;

[0012] readjusting the initial admission condition according to the second user data, and obtaining a target admission condition; and

[0013] determining a credit line of each target user that conforms to the target admission condition according to the first user data and the second user data.

[0014] Further, the step of employing a preset algorithm to analyze the first user data, and obtaining a target variable to which the target user corresponds includes:

[0015] constructing a risk variable according to the first user data and a preset rule, wherein the risk variable is used to describe credit risk of the target user; and

[0016] employing a preset algorithm to analyze the risk variable according to preset dimensions, and screening the target variable out of the risk variable according to an analyzing result and a preset screening threshold.

[0017] Further, the preset algorithm includes a first preset algorithm and a second preset algorithm, the screening threshold includes a first screening threshold and a second screening threshold, and the step of employing a preset algorithm to analyze the risk variable according to preset dimensions, and screening the target variable out of the risk variable according to an analyzing result and a preset screening threshold includes:

[0018] employing the first preset algorithm to analyze the risk variable according to preset dimensions, and obtaining a first analyzing result;

[0019] screening out a plurality of risk variables according to the first analyzing result and the preset first screening threshold;

[0020] employing the second preset algorithm to analyze the plural risk variables, and obtaining a second analyzing result; and Date recue / Date received 2021-12-09

[0021] screening the target variable out of the plural risk variables according to the second analyzing result and the preset second screening threshold.

[0022] Further, the second preset algorithm includes a regression algorithm, and the step of employing the second preset algorithm to analyze the plural risk variables, and obtaining a second analyzing result includes:

[0023] performing a discretization process on the plural risk variables, and obtaining fragments to which the plural risk variables correspond; and

[0024] employing the regression algorithm to perform regression analysis on the fragments, and obtaining the second analyzing result.

[0025] Further, the step of employing a preset algorithm to analyze the first user data, and obtaining a target variable to which the target user corresponds includes:

[0026] classifying the target user according to a preset classifying rule, and obtaining a classified type to which each target user corresponds;

[0027] determining any target user that satisfies a preset analyzing condition according to the classified type; and

[0028] employing a preset algorithm to analyze the first user data to which the target user that satisfies the analyzing condition corresponds, and obtaining a corresponding target variable.

[0029] Further, the step of determining a credit line of each target user that conforms to the target admission condition according to the first user data and the second user data includes:

[0030] determining a first line of each target user that conforms to the target admission condition and a coefficient to which the first line corresponds according to the first user data and a preset line table;

[0031] determining a second line of each target user that conforms to the target admission condition according to the second user data; and

[0032] calculating to obtain the credit line of each target user that conforms to the target admission condition according to the first line, the coefficient, and the second line.

Date recue / Date received 2021-12-09

[0033] Further, the step of obtaining a first user data of a target user on a first target platform includes:

[0034] sending a data request requesting the first user data of the target user to the first target platform, wherein the data request includes a data format of the first user data; and

[0035] receiving the first user data that conforms to the data format returned by the first target platform.

[0036] According to the second aspect, there is provided a device for assessing user credit risk, and the device comprises:

[0037] a data obtaining module, for obtaining a first user data of a target user on a first target platform and a second user data of the target user on a second target platform;

[0038] a variable obtaining module, for employing a preset algorithm to analyze the first user data, and obtaining a target variable to which the target user corresponds, wherein the target variable includes a variable describing credit risk of the target user;

[0039] a condition determining module, for determining an initial admission condition of credit according to the target variable and a preset threshold;

[0040] a condition readjusting module, for readjusting the initial admission condition according to the second user data, and obtaining a target admission condition; and

[0041] a line determining module, for determining a credit line of each target user that conforms to the target admission condition according to the first user data and the second user data.

[0042] According to the third aspect, there is provided a computer equipment that comprises a memory, a processor and a computer program stored on the memory and operable on the processor, and the following steps are realized when the processor executes the computer program:

[0043] obtaining a first user data of a target user on a first target platform and a second user data of the target user on a second target platform;

[0044] employing a preset algorithm to analyze the first user data, and obtaining a target Date recue / Date received 2021-12-09 variable to which the target user corresponds, wherein the target variable includes a variable describing credit risk of the target user;

[0045] determining an initial admission condition of credit according to the target variable and a preset threshold;

[0046] readjusting the initial admission condition according to the second user data, and obtaining a target admission condition; and

[0047] determining a credit line of each target user that conforms to the target admission condition according to the first user data and the second user data.

[0048] According to the fourth aspect, there is provided a computer-readable storage medium storing a computer program thereon, and the following steps are realized when the computer program is executed by a processor:

[0049] obtaining a first user data of a target user on a first target platform and a second user data of the target user on a second target platform;

[0050] employing a preset algorithm to analyze the first user data, and obtaining a target variable to which the target user corresponds, wherein the target variable includes a variable describing credit risk of the target user;

[0051] determining an initial admission condition of credit according to the target variable and a preset threshold;

[0052] readjusting the initial admission condition according to the second user data, and obtaining a target admission condition; and

[0053] determining a credit line of each target user that conforms to the target admission condition according to the first user data and the second user data.

[0054] The technical solutions provided by the embodiments of the present invention achieve the following advantageous effects.

[0055] In the method of and device for assessing user credit risk, computer equipment and storage medium provided by the embodiments of the present invention, by obtaining a first user Date recue / Date received 2021-12-09 data of a target user on a first target platform and a second user data of the target user on a second target platform, employing a preset algorithm to analyze the first user data, obtaining a target variable to which the target user corresponds, the target variable including a variable describing credit risk of the target user, determining an initial admission condition of credit according to the target variable and a preset threshold, readjusting the initial admission condition according to the second user data, obtaining a target admission condition, and determining a credit line of each target user that conforms to the target admission condition according to the first user data and the second user data, the present invention makes full and effective use of data provided by a first target platform (such as a core enterprise in the supply chain), sets well-directed risk variables, and hence sets corresponding credit admission conditions and credit lines to reduce bad loan rate of credit, and to reduce time consumption in collecting client information and credit cost at the same time.
BRIEF DESCRIPTION OF THE DRAWINGS

[0056] In order to more clearly describe the technical solutions in the embodiments of the present invention, drawings needed to illustrate the embodiments will be briefly introduced below. Apparently, the drawings introduced below are merely directed to some embodiments of the present invention, and persons ordinarily skilled in the art may base on these drawings to further acquire other drawings without spending creative effort in the process.

[0057] Fig. 1 is a flowchart illustrating the method of assessing user credit risk according to an exemplary embodiment;

[0058] Fig. 2 is a view schematically illustrating the structure of a device for assessing user credit risk according to an exemplary embodiment; and

[0059] Fig. 3 is a view schematically illustrating the internal structure of a computer equipment according to an exemplary embodiment.

Date recue / Date received 2021-12-09 DETAILED DESCRIPTION OF THE INVENTION

[0060] To make more clear the objectives, technical solutions and advantages of the present invention, technical solutions in the embodiments of the present invention will be described more clearly and comprehensively below with reference to the drawings accompanying the embodiments of the present invention. Apparently, the embodiments as described here are merely partial, rather than the entire, embodiments of the present invention.
All other embodiments obtainable by persons ordinarily skilled in the art on the basis of the embodiments in the present invention without spending creative effort shall all fall within the protection scope of the present invention.

[0061] As noted in the Description of Related Art, the prior-art credit risk assessing modes on the one hand cannot acquire complete client information, operational information of the client in particular, whereby bad loan rate is rendered high, on the other hand, collection of client information via prior-art information obtaining mode is problematic in terms of long time duration and high cost of manpower, etc.

[0062] In order to overcome the aforementioned problems, in the embodiments of the present invention is creatively proposed a method of assessing user credit risk, the method utilizes user information data provided by a core enterprise in the supply chain (namely a first target platform) to extract information capable of differentiating clients credits, sets admission conditions and credit lines, and performs overall judgment on clients credits in combination with data provided by a third-party servicing company (such as a credit investigation company), thereby carries out credit loans to downstream clients of the core enterprise that satisfy the conditions, whereby credit risk assessing efficiency is enhanced, and bad loan rate of credit is lowered at the same time.

[0063] Fig. 1 is a flowchart illustrating the method of assessing user credit risk according to an exemplary embodiment. With reference to Fig. 1, the method comprises the following steps.

Date recue / Date received 2021-12-09

[0064] Si - obtaining a first user data of a target user on a first target platform and a second user data of the target user on a second target platform.

[0065] Specifically, the core enterprise of the supply chain usually possesses large data of downstream clients (such as small and micro enterprises), these data are mainly the information of the downstream clients picking up (purchasing) goods from the core enterprise, including, but not limited to, times and quota of picking-up per month, seasonal fluctuation rates of goods, categories of goods, and cooperation durations with the core enterprise, and so on. As a kernel link in the supply chain, the core enterprise sometimes collects some basic information of the clients, including the locations where clients businesses concentrate, and the circumstances of the downstream clients, etc. Many core enterprises further have their own reward and punishment mechanisms with respect to clients delivery, rewarding clients to pick up goods according to targets or even over the targets, these information and the information of downstream clients having been previously overdue in other financial institutions serve as very good indicators with which to judge the credits of the downstream clients.

[0066] In the embodiments of the present invention, the target user includes, but is not limited to, a small and micro enterprise, the first target platform includes, but is not limited to, a core enterprise in the supply chain, the first user data includes user data possessed by the core enterprise, the second target platform includes, but is not limited to, a credit investigation company, and the second user data includes credit investigation data of the user.

[0067] S2 - employing a preset algorithm to analyze the first user data, and obtaining a target variable to which the target user corresponds, wherein the target variable includes a variable describing credit risk of the target user.

[0068] Specifically, after the first user data provided by the first target platform has been obtained, it is required to base on the first user data to analyze the circumstances of the target Date recue / Date received 2021-12-09 user. During specific implementation, a preset algorithm can be employed to analyze the first user data, and the analyzed contents include, but are not limited to, distribution of various features of the client, distribution of circumstances of the client whose credits have been overdue in other banks, distribution of rewards and punishments for delivery, and distribution of delivery amounts of the client, etc. A risk indicator (namely target variable) is reasonably set in conjunction with the distribution circumstances of data and overdue features.

[0069] S3 - determining an initial admission condition of credit according to the target variable and a preset threshold.

[0070] Specifically, after the target variable has been determined, a corresponding initial admission condition is set in conjunction with the target variable and a preset threshold. Besides the target variable as chosen, consideration can be also given to some apparently financially meaningful conditions. Setup of the initial admission condition here can be combined with the analyzing result in the previous step.

[0071] The initial admission condition includes, but is not limited to, the following.

[0072] no appearance on the blacklist of the core enterprise of the supply chain;

[0073] exclusion of clients with inferior or severe overdue circumstance (over 7 days) on the credit list of the core enterprise of the supply chain;

[0074] exclusion of provinces or cities to which the business is temporarily not extended (due to reasons of small business quantity, undeveloped economy or inferior credit environment, etc.);

[0075] exclusion of clients whose yearly picking-ups of goods are less than three months (low picking-ups indicate unstable business);

[0076] exclusion of clients who pick up goods in just one month or less than one month in the recent three months;

[0077] exclusion of clients whose amount of goods picked up is less than 500,000 yuan RMB
(unduly low picking-up quota cannot get high credit line, being low in risk but also low in efficiency);
Date recue / Date received 2021-12-09

[0078] exclusion of clients whose order completion rate is lower than 30%;

[0079] exclusion of clients whose amount of goods picked up is reduced over 50% on a year-on-year or month-on-month basis within one year; and

[0080] exclusion of clients who opened accounts at the core enterprise less than one year.

[0081] S4 - readjusting the initial admission condition according to the second user data, and obtaining a target admission condition.

[0082] Specifically, the third-party credit investigation company and the Central Bank can also provide a lot of information relevant to a downstream client (a target user), such as business information of the client, judicially involved information, social loan information, etc. These information can provide fuller knowledge of each downstream client. The information provided by the core enterprise (first user data) and the credit investigation information (second user information) are combined to readjust the initial admission condition obtained in the previous step to obtain target admission data.

[0083] S5 - determining a credit line of each target user that conforms to the target admission condition according to the first user data and the second user data.

[0084] Specifically, by analyzing the first user data obtained from the first target platform in combination with experiences acquired from other projects can be sorted out the upper limit of creditable line of each target user. On the other hand, from information provided by the Central Bank or a third-party credit investigation company (namely the second user data of the second target platform) can be analyzed out the credit circumstances of each target user in other banks, including creditable line, already used line, overdue or collection information, etc. The actual creditable line of each target user can be obtained by combining the information of these two aspects.

[0085] As a preferred mode of execution in the embodiments of the present invention, the step Date recue / Date received 2021-12-09 of employing a preset algorithm to analyze the first user data, and obtaining a target variable to which the target user corresponds includes:

[0086] constructing a risk variable according to the first user data and a preset rule, wherein the risk variable is used to describe credit risk of the target user; and

[0087] employing a preset algorithm to analyze the risk variable according to preset dimensions, and screening the target variable out of the risk variable according to an analyzing result and a preset screening threshold.

[0088] Specifically, constructing a risk variable is to construct those potential variables that are capable of reflecting credit levels of the user. The risk variable constructed according to the first user data and a preset rule includes, but is not limited to, the number of months within a year in which orders are completed, goods picking-up quota in one year, credit balance, the number of months in which goods are picked up in the last year, the number of months in which goods are picked up in the last six months, the number of months in which goods are picked up in the last three months, goods picking-up quota in the last year, goods picking-up quota in the last six months, goods picking-up quota in the last three months, year-on-year growth rate of goods picked up in the last year, month-on-month growth rate of goods picked up in the last year, year-on-year growth rate of goods picked up in the last six months, month-on-month growth rate of goods picked up in the last six months, year-on-year growth rate of goods picked up in the last three months, month-on-month growth rate of goods picked up in the last three months, the region (province, city) where the client resides, the time when the client opened account, the amount of orders completed in the last year, etc. Economically speaking, these variables are related in certain extent to the non-performance of the client.

[0089] When the preset algorithm is employed to analyze the risk variable, dimension information to be analyzed can be set in advance according to actual requirements, the analyzing result as obtained is compared with a preset screening threshold, and a target variable that conforms to the requirements is screened out of the risk variable.

Date recue / Date received 2021-12-09

[0090] As a preferred mode of execution in the embodiments of the present invention, the preset algorithm includes a first preset algorithm and a second preset algorithm, the screening threshold includes a first screening threshold and a second screening threshold, and the step of employing a preset algorithm to analyze the risk variable according to preset dimensions, and screening the target variable out of the risk variable according to an analyzing result and a preset screening threshold includes:

[0091] employing the first preset algorithm to analyze the risk variable according to preset dimensions, and obtaining a first analyzing result;

[0092] screening out a plurality of risk variables according to the first analyzing result and the preset first screening threshold;

[0093] employing the second preset algorithm to analyze the plural risk variables, and obtaining a second analyzing result; and

[0094] screening the target variable out of the plural risk variables according to the second analyzing result and the preset second screening threshold.

[0095] Specifically, in the embodiments of the present invention, employing the preset algorithm to analyze the risk variable according to preset dimensions includes analyzing a single risk variable and analyzing a plurality of risk variables, and employing the first preset algorithm to analyze the risk variable according to preset dimensions includes, but is not limited to, the following:

[0096] analyzing the number of months within one year in which the target user completes orders, the distribution of months, and the relation between goods picking-up quota in a total year and non-performance of the client ¨ usually speaking, the more number of months within one year in which orders are completed, the better;

[0097] analyzing credit balance distribution, overdue or non-performing balance distribution of the target user, and the relation between credit balance and overdue or non-performance on the basis of the first user data provided by the first target platform ¨ severely overdue (such as over Date recue / Date received 2021-12-09 30 days) and non-performing clients should be avoided;

[0098] analyzing the number of months in the last year in which goods are picked up, the number of months in the last six months in which goods are picked up, and the number of months in the last three months in which goods are picked up by each target user, and analyzing the relation between these data and change in goods picking-up quota of the client and possible overdue circumstances ¨ the more number of months in which goods are picked up, the better is the downstream business;

[0099] analyzing the relation between non-performance and the year-on-year and month-on-month growth rates of goods picked up in the last year, the year-on-year and month-on-month growth rates of goods picked up in the last six months, and the year-on-year and month-on-month growth rates of goods picked up in the last three months, and analyzing the relation between these growth rates and change in goods picking-up quota of the client in a total year and possible overdue circumstances ¨ usually speaking, a well operating company achieves stable increase in both the year-on-year and month-on-month growth rates of goods picking-up quota, and there is better not much fluctuation, as a less fluctuating and stably increasing client is the first choice for credits;

[0100] analyzing the relation between the region where the target user resides and the non-performance of the client ¨ different regions have different economic development levels, credit environments and credit cultures, clients from regions with higher economic development levels and better credit cultures are the first choice, and region selection should take potential business volumes, credit environments and economic development levels of regions into overall consideration;

[0101] analyzing the relation between the time when the target user opened account and the non-performance of the client ¨ in the case the account was opened too short, there is no sufficient historical data accumulation of the client, and there is also no way to analyze Date recue / Date received 2021-12-09 corresponding data and behaviors; generally speaking, the client's account should have been opened at least one year before sufficient goods picking-up history of the client can be accumulated for analysis;

[0102] analyzing the relation between the non-performance of the client and the goods picking-up quota in the last year, the goods picking-up quota in the last sixth months, and the goods picking-up quota in the last three months of the target user ¨ the goods picking-up quota is the data upon which the credit line mainly depends, the higher and more stable the goods picking-up quota of the client is, the better it will be, and the change in the goods picking-up quota of the client is closely related to the change in the operational circumstances of this client, so it is required to clearly analyze the potential tendency of the goods picking-up quota; and

[0103] analyzing the relation between the order completion rate of the target user in the core enterprise and the non-performance of the client ¨ the levels whereby orders are completed predict to some extent the levels of operations of the downstream enterprises.

[0104] The relations analyzed above can be reflected through certain statistics, including such statistic indices as AUC and KS. A first screening threshold is provided to the statistic index of each risk variable, and risk variables exceeding the certain first screening threshold are preliminarily selected to pass. After passing the first round of screening, plural risk variables possessing the capability to apparently differentiating client risks are screened out to serve as basis for the next round of analyzing a plurality of risk variables.

[0105] Analyzing a plurality of risk variables mainly aims to find out a collection of variables most capable of differentiating client credit levels. Likewise, a second screening threshold can be preset while the plural risk variables are being analyzed, the analyzing result is compared with the second screening threshold, and the optimal variable is screened out of the plural risk variables to serve as the target variable.
Date recue / Date received 2021-12-09

[0106] As a preferred mode of execution in the embodiments of the present invention, the second preset algorithm includes a regression algorithm, and the step of employing the second preset algorithm to analyze the plural risk variables, and obtaining a second analyzing result includes:

[0107] performing a discretization process on the plural risk variables, and obtaining fragments to which the plural risk variables correspond; and

[0108] employing the regression algorithm to perform regression analysis on the fragments, and obtaining the second analyzing result.

[0109] Specifically, many methods can be employed to analyze the plural risk variables, such as methods of linear regression, logistic regression and decision tree. Taking logistic regression for example, the logistic regression can employ the modes of forward regression, backward regression, and simultaneous forward regression and backward regression. What a mode is specifically employed can be decided by comparing plural results to see which mode is the best.

[0110] Before the plural risk variables are analyzed, the variables should be converted, and such conversion includes, but is not limited to, the WOE mode, namely to subject the variables to discretization to be partitioned into plural fragments, and to calculate the ODDS of each fragment, namely a ratio between quality clients and inferior clients, which ratio is used to enter regression analysis of the next step. Judging which variables are the best can be measured by indicators with regression effects, and the indicators with regression effects include such statistics as AUC value, KS value and relevancy, etc., to which no definition is made in the embodiments of the present invention, as it is possible for users to base on actual requirements to select one or more therefrom. Taking the AUC value for example, a collection of variables with the best AUC value is selected through the foregoing steps to serve as the target variable.

[0111] As a preferred mode of execution in the embodiments of the present invention, the coefficient of the regression result of the aforementioned plural risk variables can be further taken as a reference to setting the initial admission condition. When the product of the coefficient Date recue / Date received 2021-12-09 with the ODDS of a corresponding variable is apparently lower than the average value, for instance, lower than twice as much as the average value minus a standard difference, then the value of a variable to which the value twice as much as the average value minus a standard difference corresponds can be set as a threshold. Each variable is performed with similar analysis before the corresponding threshold of each variable can be obtained.

[0112] As a preferred mode of execution in the embodiments of the present invention, the step of employing a preset algorithm to analyze the first user data, and obtaining a target variable to which the target user corresponds includes:

[0113] classifying the target user according to a preset classifying rule, and obtaining a classified type to which each target user corresponds;

[0114] determining any target user that satisfies a preset analyzing condition according to the classified type; and

[0115] employing a preset algorithm to analyze the first user data to which the target user that satisfies the analyzing condition corresponds, and obtaining a corresponding target variable.

[0116] Specifically, in order to enhance risk assessing precision, in the embodiments of the present invention, when the preset algorithm is employed to analyze the first user data, the first user data of a non-performing client (also referred to as "inferior client") can be selected for analysis, so as to analyze out features of the non-performing client to serve as reference to subsequent setting of the admission condition. Accordingly, it is possible to firstly ascertain definitions of the non-performing clients; during specific implementation, clients whose loans were once overdue, clients appearing on the blacklist of the core enterprise, and clients having no amount paid to picked up goods in six months are defined as inferior clients or non-performing clients, while clients clear of the above circumstances are defined as quality clients.

[0117] Target users are classified according to the above definitions, the classified type to which each target user corresponds is judged as to whether the target user is a quality client or an inferior client, those target users that pertain to non-performing clients are screened out, and the Date recue / Date received 2021-12-09 first user data to which these target users correspond is analyzed to obtain corresponding target variables.

[0118] As a preferred mode of execution in the embodiments of the present invention, the step of determining a credit line of each target user that conforms to the target admission condition according to the first user data and the second user data includes:

[0119] determining a first line of each target user that conforms to the target admission condition and a coefficient to which the first line corresponds according to the first user data and a preset line table;

[0120] determining a second line of each target user that conforms to the target admission condition according to the second user data; and

[0121] calculating to obtain the credit line of each target user that conforms to the target admission condition according to the first line, the coefficient, and the second line.

[0122] Specifically, in the embodiments of the present invention, it is possible to base on actual requirements to set the formula for calculating the credit line of a target user, as a preferred example, the calculation formula can be as shown below:
credit line = max (basic line * coefficient ¨ Central Bank credit investigation used line, 0)

[0123] What should be determined in the formula are the two parameters of basic line and coefficient, the Central Bank credit investigation used line indicates credit extension obtained in the name of an enterprise, calculated is only the used credit line already extracted, and this can be extracted from a credit investigation report of the Central Bank. The basic line is the first line, and the Central Bank credit investigation used line is the second line. The principle of setting credit lines of the target user is that the profitability of the client is capable of covering the existent credit line and the newly added credit line.

[0124] The basic line mainly takes into consideration the safety of the line as provided, after the user has passed the target admission condition, the basic line is mainly related to the Date recue / Date received 2021-12-09 repayment capability of the user, while the repayment capability is mainly related to the income of the user. By way of example, the sales amount of money (goods picking-up amount of money) of the user can be used as a benchmark for consideration, and the basic line is realized by multiplying the benchmark with a certain proportion. A reasonable base is 20%
to 40% of the goods picking-up amount of money, the consideration here is that a certain proportion of the goods picking-up amount of money should be in loan, and the loan limit is risky when this certain proportion is exceeded; after a certain period of time with credit history in the future, this proportion may be increased or decreased according to non-performing circumstances of the client. This proportion is also closely related to the core enterprise (first target platform), as the stronger the core enterprise is, the higher can be the proportion.

[0125] As an example, calculation of the basic line can be as shown in the following Table.
Table 1 Table of Correspondence between Sales Amount of Money and Basic Line Sales Amount of Money of the Target Corresponding Basic Line User in the Last 12 Months 0-500,000 yuan RMB 0 yuan RMB
500,000¨one million yuan RMB 200,000 yuan RMB
one million-2.5 million yuan RMB 400,000 yuan RMB
2.5 million-5 million yuan RMB 1 million yuan RMB
million-7.5 million yuan RMB 2 million yuan RMB
7 million-10 million yuan RMB 3 million yuan RMB
over 10 million yuan RMB credit to be manually examined and approved

[0126] The corresponding coefficient should be further determined after the basic line has been well determined. The coefficient is a parameter with which to adequately readjust the basic line based on other information or newly appearing information after the basic line has been given.
The higher the coefficient is, the better will be the credit of the enterprise, otherwise, a lower Date recue / Date received 2021-12-09 coefficient indicates that the enterprise falls short of perfect credit or leaves something to be desired of the credit.

[0127] The coefficient is preliminarily determined as follows:
Table 2 Table of Correspondence between Order Completion Circumstance and Coefficient Order Completion Circumstance in the Corresponding Coefficient Last 12 Months 80%+ order overfulfilled 1 65%-80% order overfulfilled 0.9 50%-65% order overfulfilled 0.8 35%-50% order overfulfilled 0.7 0-35% order overfulfilled over upper limit 0.7, to be decided by credit examining personnel

[0128] After the coefficient has been determined, it is then considered to readjust the coefficient.
Readjustment of the coefficient mainly reflects the information of the user relevant to recent change in credit risk, and the information includes possible negative early warning signals, abrupt decrease in the sales (goods picking-up) amount of money, and overdue circumstances of currently existing credits, etc. And readjustment of the coefficient aims to adequately lower the credit level of the client, control the credit line of the client, and even reject loan to the client when potential risk is encountered.

[0129] The main conditions relevant to readjustment of the coefficient include, but are not limited to, the following:

[0130] lowering the coefficient for one grade if there is an early warning signal against the target user;

[0131] lowering the coefficient for one grade if the goods picking-up amount of money of the target user decreases by 20% in the recent three months on a year-on-year basis;

[0132] lowering the coefficient for one grade if the goods picking-up amount of money of the Date recue / Date received 2021-12-09 target user decreases by 15% in last year on a year-on-year basis;

[0133] lowering the coefficient for one grade if abnormality occurs to the target user; and

[0134] lowering the coefficient for one grade if the target user had overdue circumstances in the last two years, with the maximum overdue period being less than 15 days, and all overdue loans have been made good.

[0135] As a preferred mode of execution in the embodiments of the present invention, the step of obtaining a first user data of a target user on a first target platform includes:

[0136] sending a data request requesting the first user data of the target user to the first target platform, wherein the data request includes a data format of the first user data; and

[0137] receiving the first user data that conforms to the data format returned by the first target platform.

[0138] Specifically, the core enterprise usually possesses a lot of information of downstream clients, in order to enhance availability of the data, when the first user data of the target user is requested from the first target platform, such contents as the format of the required data and time duration (over two years in general) of the required data can be ascertained.

[0139] Fig. 2 is a view schematically illustrating the structure of a device for assessing user credit risk according to an exemplary embodiment. The device comprises:

[0140] a data obtaining module, for obtaining a first user data of a target user on a first target platform and a second user data of the target user on a second target platform;

[0141] a variable obtaining module, for employing a preset algorithm to analyze the first user data, and obtaining a target variable to which the target user corresponds, wherein the target variable includes a variable describing credit risk of the target user;

[0142] a condition determining module, for determining an initial admission condition of credit according to the target variable and a preset threshold;

[0143] a condition readjusting module, for readjusting the initial admission condition according to the second user data, and obtaining a target admission condition; and Date recue / Date received 2021-12-09

[0144] a line determining module, for determining a credit line of each target user that conforms to the target admission condition according to the first user data and the second user data.

[0145] As a preferred mode of execution in the embodiments of the present invention, the variable obtaining module is specifically employed for:

[0146] constructing a risk variable according to the first user data and a preset rule, wherein the risk variable is used to describe credit risk of the target user; and

[0147] employing a preset algorithm to analyze the risk variable according to preset dimensions, and screening the target variable out of the risk variable according to an analyzing result and a preset screening threshold.

[0148] As a preferred mode of execution in the embodiments of the present invention, the variable obtaining module is specifically employed for:

[0149] employing the first preset algorithm to analyze the risk variable according to preset dimensions, and obtaining a first analyzing result;

[0150] screening out a plurality of risk variables according to the first analyzing result and the preset first screening threshold;

[0151] employing the second preset algorithm to analyze the plural risk variables, and obtaining a second analyzing result; and

[0152] screening the target variable out of the plural risk variables according to the second analyzing result and the preset second screening threshold.

[0153] As a preferred mode of execution in the embodiments of the present invention, the variable obtaining module is further employed for:

[0154] performing a discretization process on the plural risk variables, and obtaining fragments to which the plural risk variables correspond; and

[0155] employing the regression algorithm to perform regression analysis on the fragments, and obtaining the second analyzing result.

Date recue / Date received 2021-12-09

[0156] As a preferred mode of execution in the embodiments of the present invention, the variable obtaining module is further employed for:

[0157] classifying the target user according to a preset classifying rule, and obtaining a classified type to which each target user corresponds;

[0158] determining any target user that satisfies a preset analyzing condition according to the classified type; and

[0159] employing a preset algorithm to analyze the first user data to which the target user that satisfies the analyzing condition corresponds, and obtaining a corresponding target variable.

[0160] As a preferred mode of execution in the embodiments of the present invention, the line determining module is specifically employed for:

[0161] determining a first line of each target user that conforms to the target admission condition and a coefficient to which the first line corresponds according to the first user data and a preset line table;

[0162] determining a second line of each target user that conforms to the target admission condition according to the second user data; and

[0163] calculating to obtain the credit line of each target user that conforms to the target admission condition according to the first line, the coefficient, and the second line.

[0164] As a preferred mode of execution in the embodiments of the present invention, the data obtaining module is specifically employed for:

[0165] sending a data request requesting the first user data of the target user to the first target platform, wherein the data request includes a data format of the first user data; and

[0166] receiving the first user data that conforms to the data format returned by the first target platform.

[0167] Fig. 3 is a view schematically illustrating the internal structure of a computer equipment according to an exemplary embodiment. With reference to Fig. 3, the computer equipment comprises a processor, a memory, and a network interface connected to each other via a system Date recue / Date received 2021-12-09 bus. The processor of the computer equipment is employed to provide computing and controlling capabilities. The memory of the computer equipment includes a nonvolatile storage medium, and an internal memory. The nonvolatile storage medium stores therein an operating system, a computer program and a database. The internal memory provides environment for the running of the operating system and the computer program in the nonvolatile storage medium. The network interface of the computer equipment is employed to connect to an external terminal via network for communication. The computer program realizes a method of optimizing an execution plan when it is executed by a processor.

[0168] As understandable to persons skilled in the art, the structure illustrated in Fig. 3 is merely a block diagram of partial structure relevant to the solution of the present invention, and does not constitute any restriction to the computer equipment on which the solution of the present invention is applied, as the specific computer equipment may comprise component parts that are more than or less than those illustrated in Fig. 3, or may combine certain component parts, or may have different layout of component parts.

[0169] As a preferred mode of execution in the embodiments of the present invention, there is provided a computer equipment that comprises a memory, a processor and a computer program stored on the memory and operable on the processor, and the following steps are realized when the processor executes the computer program:

[0170] obtaining a first user data of a target user on a first target platform and a second user data of the target user on a second target platform;

[0171] employing a preset algorithm to analyze the first user data, and obtaining a target variable to which the target user corresponds, wherein the target variable includes a variable describing credit risk of the target user;

[0172] determining an initial admission condition of credit according to the target variable and a preset threshold;

[0173] readjusting the initial admission condition according to the second user data, and obtaining a target admission condition; and Date recue / Date received 2021-12-09

[0174] determining a credit line of each target user that conforms to the target admission condition according to the first user data and the second user data.

[0175] As a preferred mode of execution in the embodiments of the present invention, the following steps are further realized when the processor executes the computer program:

[0176] constructing a risk variable according to the first user data and a preset rule, wherein the risk variable is used to describe credit risk of the target user; and

[0177] employing a preset algorithm to analyze the risk variable according to preset dimensions, and screening the target variable out of the risk variable according to an analyzing result and a preset screening threshold.

[0178] As a preferred mode of execution in the embodiments of the present invention, the following steps are further realized when the processor executes the computer program:

[0179] employing the first preset algorithm to analyze the risk variable according to preset dimensions, and obtaining a first analyzing result;

[0180] screening out a plurality of risk variables according to the first analyzing result and the preset first screening threshold;

[0181] employing the second preset algorithm to analyze the plural risk variables, and obtaining a second analyzing result; and

[0182] screening the target variable out of the plural risk variables according to the second analyzing result and the preset second screening threshold.

[0183] As a preferred mode of execution in the embodiments of the present invention, the following steps are further realized when the processor executes the computer program:

[0184] performing a discretization process on the plural risk variables, and obtaining fragments to which the plural risk variables correspond; and

[0185] employing the regression algorithm to perform regression analysis on the fragments, and obtaining the second analyzing result.
Date recue / Date received 2021-12-09

[0186] As a preferred mode of execution in the embodiments of the present invention, the following steps are further realized when the processor executes the computer program:

[0187] classifying the target user according to a preset classifying rule, and obtaining a classified type to which each target user corresponds;

[0188] determining any target user that satisfies a preset analyzing condition according to the classified type; and

[0189] employing a preset algorithm to analyze the first user data to which the target user that satisfies the analyzing condition corresponds, and obtaining a corresponding target variable.

[0190] As a preferred mode of execution in the embodiments of the present invention, the following steps are further realized when the processor executes the computer program:

[0191] determining a first line of each target user that conforms to the target admission condition and a coefficient to which the first line corresponds according to the first user data and a preset line table;

[0192] determining a second line of each target user that conforms to the target admission condition according to the second user data; and

[0193] calculating to obtain the credit line of each target user that conforms to the target admission condition according to the first line, the coefficient, and the second line.

[0194] As a preferred mode of execution in the embodiments of the present invention, the following steps are further realized when the processor executes the computer program:

[0195] sending a data request requesting the first user data of the target user to the first target platform, wherein the data request includes a data format of the first user data; and

[0196] receiving the first user data that conforms to the data format returned by the first target platform.

[0197] In the embodiments of the present invention, there is further provided a computer-readable storage medium storing thereon a computer program, and the following steps are realized when the computer program is realized by a processor:

Date recue / Date received 2021-12-09

[0198] obtaining a first user data of a target user on a first target platform and a second user data of the target user on a second target platform;

[0199] employing a preset algorithm to analyze the first user data, and obtaining a target variable to which the target user corresponds, wherein the target variable includes a variable describing credit risk of the target user;

[0200] determining an initial admission condition of credit according to the target variable and a preset threshold;

[0201] readjusting the initial admission condition according to the second user data, and obtaining a target admission condition; and

[0202] determining a credit line of each target user that conforms to the target admission condition according to the first user data and the second user data.

[0203] As a preferred mode of execution in the embodiments of the present invention, the following steps are further realized when the computer program is executed by a processor:

[0204] constructing a risk variable according to the first user data and a preset rule, wherein the risk variable is used to describe credit risk of the target user; and

[0205] employing a preset algorithm to analyze the risk variable according to preset dimensions, and screening the target variable out of the risk variable according to an analyzing result and a preset screening threshold.

[0206] As a preferred mode of execution in the embodiments of the present invention, the following steps are further realized when the computer program is executed by a processor:

[0207] employing the first preset algorithm to analyze the risk variable according to preset dimensions, and obtaining a first analyzing result;

[0208] screening out a plurality of risk variables according to the first analyzing result and the preset first screening threshold;

[0209] employing the second preset algorithm to analyze the plural risk variables, and obtaining a second analyzing result; and

[0210] screening the target variable out of the plural risk variables according to the second Date recue / Date received 2021-12-09 analyzing result and the preset second screening threshold.

[0211] As a preferred mode of execution in the embodiments of the present invention, the following steps are further realized when the computer program is executed by a processor:

[0212] performing a discretization process on the plural risk variables, and obtaining fragments to which the plural risk variables correspond; and

[0213] employing the regression algorithm to perform regression analysis on the fragments, and obtaining the second analyzing result.

[0214] As a preferred mode of execution in the embodiments of the present invention, the following steps are further realized when the computer program is executed by a processor:

[0215] classifying the target user according to a preset classifying rule, and obtaining a classified type to which each target user corresponds;

[0216] determining any target user that satisfies a preset analyzing condition according to the classified type; and

[0217] employing a preset algorithm to analyze the first user data to which the target user that satisfies the analyzing condition corresponds, and obtaining a corresponding target variable.

[0218] As a preferred mode of execution in the embodiments of the present invention, the following steps are further realized when the computer program is executed by a processor:

[0219] determining a first line of each target user that conforms to the target admission condition and a coefficient to which the first line corresponds according to the first user data and a preset line table;

[0220] determining a second line of each target user that conforms to the target admission condition according to the second user data; and

[0221] calculating to obtain the credit line of each target user that conforms to the target admission condition according to the first line, the coefficient, and the second line.

[0222] As a preferred mode of execution in the embodiments of the present invention, the Date recue / Date received 2021-12-09 following steps are further realized when the computer program is executed by a processor:

[0223] sending a data request requesting the first user data of the target user to the first target platform, wherein the data request includes a data format of the first user data; and

[0224] receiving the first user data that conforms to the data format returned by the first target platform.

[0225] In summary, the technical solutions provided by the embodiments of the present invention achieve the following advantageous effects.

[0226] In the method of and device for assessing user credit risk, computer equipment and storage medium provided by the embodiments of the present invention, by obtaining a first user data of a target user on a first target platform and a second user data of the target user on a second target platform, employing a preset algorithm to analyze the first user data, obtaining a target variable to which the target user corresponds, the target variable including a variable describing credit risk of the target user, determining an initial admission condition of credit according to the target variable and a preset threshold, readjusting the initial admission condition according to the second user data, obtaining a target admission condition, and determining a credit line of each target user that conforms to the target admission condition according to the first user data and the second user data, the present invention makes full and effective use of data provided by a first target platform (such as a core enterprise in the supply chain), sets well-directed risk variables, and hence sets corresponding credit admission conditions and credit lines to reduce bad loan rate of credit, and to reduce time consumption in collecting client information and credit cost at the same time.

[0227] As should be noted, when the device for assessing user credit risk provided by the aforementioned embodiments triggers the risk assessing business, it is only exemplarily described with the division into the aforementioned various functional modules, while in practical application, it is possible to assign the functions to different functional modules to be completed thereby according to requirements, that is to say, to divide the internal structure of Date recue / Date received 2021-12-09 the device into different functional modules to complete the entire or partial functions described above. In addition, the device for assessing user credit risk provided by the aforementioned embodiments pertains to the same conception as the method of assessing user credit risk, that is to say, the device is based on the method of assessing user credit risk ¨
details for its specific realization process can be inferred from the method embodiments, while no repetition would be made in this context.

[0228] As comprehensible to persons ordinarily skilled in the art, the entire or partial steps of the aforementioned embodiments can be completed via hardware, or via a program instructing relevant hardware, the program can be stored in a computer-readable storage medium, and the storage medium can be a read-only memory, a magnetic disk, or an optical disk.

[0229] What is described above is merely directed to preferred embodiments of the present invention, and is not meant to restrict the present invention. Any amendment, equivalent replacement and improvement makeable within the spirit and scope of the present invention shall all be covered within the protection scope of the present invention.
Date recue / Date received 2021-12-09

Claims

Claims:

1. A device for assessing user credit risk, the device comprising:
a data obtaining module, for obtaining a first user data of a target user on a first target platform and a second user data of the target user on a second target platform;
a variable obtaining module for:
employing a preset algorithm to analyze the first user data;
obtaining a target variable to which the target user corresponds, wherein the target variable includes a variable describing credit risk of the target user;
a condition determining module, for determining an initial admission condition of credit according to the target variable and a preset threshold;
a condition readjusting module for:
readjusting the initial admission condition according to the second user data;
obtaining a target admission condition; and a line determining module, for determining a credit line of each target user that conforms to the target admission condition according to the first user data and the second user data.

2. The device of claim 1, wherein the variable obtaining module is further for:
constructing a risk variable according to the first user data and a preset rule, wherein the risk variable is used to describe credit risk of the target user;
employing a preset algorithm to analyze the risk variable according to preset dimensions;
and screening the target variable out of the risk variable according to an analyzing result and a preset screening threshold.

Date recue/ date received 2022-02-18

3. The device of any one of claims 1 to 2, wherein the variable obtaining module is further for:
employing the first preset algorithm to analyze the risk variable according to preset dimensions;
obtaining a first analyzing result;
screening out a plurality of risk variables according to the first analyzing result and the preset first screening threshold;
employing the second preset algorithm to analyze the plural risk variables;
obtaining a second analyzing result; and screening the target variable out of the plural risk variables according to the second analyzing result and the preset second screening threshold.

4. The device of any one of claims 1 to 3, wherein the variable obtaining module is further for:
performing a discretization process on the plural risk variables;
obtaining fragments to which the plural risk variables correspond;
employing the regression algorithm to perform regression analysis on the fragments; and obtaining the second analyzing result.

5. The device of any one of claims 1 to 4, wherein the variable obtaining module is further for:
classifying the target user according to a preset classifying rule;
obtaining a classified type to which each target user corresponds;
determining any target user that satisfies a preset analyzing condition according to the classified type;

Date recue/ date received 2022-02-18 employing the preset algorithm to analyze the first user data to which the target user that satisfies the analyzing condition corresponds; and obtaining a corresponding target variable.

6. The device of claim 1, wherein the line determining module is further for:
determining a first line of each target user that conforms to the target admission condition and a coefficient to which the first line corresponds according to the first user data and a preset line table;
determining a second line of each target user that conforms to the target admission condition according to the second user data; and calculating to obtain the credit line of each target user that conforms to the target admission condition according to the first line, the coefficient, and the second line.

7. The device of claim 1, wherein the data obtaining module is further for:
sending a data request requesting the first user data of the target user to the first target platform, wherein the data request includes a data format of the first user data; and receiving the first user data that conforms to the data format returned by the first target platform.

Date recue/ date received 2022-02-18

8. The device of any one of claims 1 to 7, wherein the target user includes any one or more of a small and micro enterprise, wherein the first target platform includes a core enterprise in a supply chain, wherein the core enterprise in the supply chain, includes any one or more of times and quota of picking-up per month, seasonal fluctuation rates of goods, categories of goods, and cooperation durations with the core enterprise, wherein the first user data includes target user data possessed by the core enterprise, wherein target user data possessed by the core enterprise includes any one or more of the locations where target users businesses concentrate, and the circumstances of downstream target users, wherein the second target platform includes a credit investigation company, and wherein the second user data includes credit investigation data of the target user.

9. The device of any one of claims 1 to 8, wherein the first user data provided by the first target platform has been obtained, based on the first user data to analyze the circumstances of the target user, wherein the preset algorithm can be employed to analyze the first user data, and the analyzed contents includes any one or more of distribution of various features of the target user, distribution of circumstances of the target user whose credits have been overdue in other banks, distribution of rewards and punishments for delivery, and distribution of delivery amounts of the target user, wherein the target variable is reasonably set in conjunction with the distribution circumstances of data and overdue features.

10. The device of any one of claims 1 to 9, wherein the target variable has been determined, the initial admission condition is set in conjunction with the target variable and the preset threshold, wherein setup of the initial admission condition here is combined with the analyzing result.

Date recue/ date received 2022-02-18

11. The device of any one of claims 1 to 10, wherein the initial admission condition includes any one or more of no appearance on a blacklist of the core enterprise of the supply chain, exclusion of target users with inferior or severe overdue circumstance, over 7 days, on a credit list of the core enterprise of the supply chain, exclusion of provinces or cities to which a business is temporarily not extended, due to reasons of small business quantity, undeveloped economy or inferior credit environment, exclusion of target users whose yearly picking-ups of goods are less than three months, exclusion of target users who pick up goods in just one month or less than one month in the recent three months, exclusion of target users whose amount of goods picked up is low, exclusion of target users whose order completion rate is lower than 30%, exclusion of target users whose amount of goods picked up is reduced over 50% on a year-on-year or month-on-month basis within one year, and exclusion of target users who opened accounts at the core enterprise less than one year.

12. The device of any one of claims 1 to 11, wherein third-party credit investigation company and a central bank provides information relevant to the downstream target user, including one or more of business information of the target user, judicially involved information, social loan information, wherein the information provided by the core enterprise and the credit investigation information are combined to readjust the initial admission condition obtained to obtain target admission data.

13. The device of any one of claims 1 to 12, wherein by analyzing the first user data obtained from the first target platform in combination with experiences acquired from other projects can be sorted out a upper limit of creditable line of each target user, wherein from information provided by the Central Bank or the third-party credit investigation company, the second user data of the second target platform, can be analyzed out credit circumstances of each target user in other banks, including one or more of creditable line, already used line, overdue or collection information, wherein the actual creditable line of each target user can be obtained by combining information of these two aspects.
Date recue/ date received 2022-02-18

14. The device of any one of claims 1 to 13, wherein constructing the risk variable is to construct those potential variables that are capable of reflecting credit levels of the target user, wherein the risk variable constructed according to the first user data and the preset rule includes one or more of number of months within a year in which orders are completed, goods picking-up quota in one year, credit balance, number of months in which goods are picked up in the last year, number of months in which goods are picked up in the last six months, number of months in which goods are picked up in last three months, goods picking-up quota in last year, goods picking-up quota in last six months, goods picking-up quota in last three months, year-on-year growth rate of goods picked up in last year, month-on-month growth rate of goods picked up in the last year, year-on-year growth rate of goods picked up in the last six months, month-on-month growth rate of goods picked up in the last six months, year-on-year growth rate of goods picked up in the last three months, month-on-month growth rate of goods picked up in the last three months, region, province, city where the target user resides, time when the target user opened account, the amount of orders completed in the last year.

15. The device of any one of claims 1 to 14, wherein the preset algorithm is employed to analyze the risk variable, dimension information to be analyzed can be set in advance according to actual requirements, wherein the analyzing result as obtained is compared with the preset screening threshold, and wherein the target variable that conforms to the requirements is screened out of the risk variable.

16. The device of any one of claims 1 to 15, wherein employing the preset algorithm to analyze the risk variable according to preset dimensions includes analyzing a single risk variable and analyzing a plurality of risk variables, and employing the first preset algorithm to analyze the risk variable according to preset dimensions comprising:
analyzing the number of months within one year in which the target user completes orders, distribution of months, and relation between goods picking-up quota in a total year and non-performance of the target user;

Date recue/ date received 2022-02-18 analyzing credit balance distribution, overdue or non-performing balance distribution of the target user, and relation between credit balance and overdue or non-performance on the basis of the first user data provided by the first target platform, wherein severely overdue and non-perfomiing users are avoided;
analyzing number of months in the last year in which goods are picked up, number of months in the last six months in which goods are picked up, and number of months in the last three months in which goods are picked up by each target user, and analyzing relation between these data and change in goods picking-up quota of the target user and possible overdue circumstances;
analyzing relation between non-performance and year-on-year and month-on-month growth rates of goods picked up in the last year, the year-on-year and month-on-month growth rates of goods picked up in the last six months, and the year-on-year and month-on-month growth rates of goods picked up in the last three months and analyzing the relation between the growth rates and change in goods picking-up quota of the target user in a total year and possible overdue circumstances;
analyzing the relation between the region where target user resides and the non-performance of the target user, wherein region selection include one or more of potential business volumes, credit environments and economic development levels of the region;
analyzing the relation between time when the target user opened account and the non-performance of target user, wherein target user's account is opened at least one year before sufficient goods picking-up history of the target user can be accumulated for analysis;
analyzing relation between the non-performance of the target user and the goods picking-up quota in the last year, the goods picking-up quota in the last sixth months, and the goods picking-up quota in the last three months of the target user; and Date recue/ date received 2022-02-18 analyzing the relation between the order completion rate of the target user in the core enterprise and the non-performance of the target user.

17. The device of any one of claims 1 to 16, wherein the relations analyzed are reflected through statistics, including statistic indices AUC and KS, wherein the first screening threshold is provided to the statistic index of each risk variable, and the risk variables exceeding the first screening threshold are preliminarily selected to pass, wherein passing the first round of screening, plural risk variables possessing capability to apparently differentiating target user risks are screened out to serve as basis for analyzing the plurality of risk variables.

18. The device of any one of claims 1 to 17, wherein analyzing the plurality of risk variables is to find out a collection of variables most capable of differentiating target user credit levels, wherein the second screening threshold can be preset while the plural risk variables are being analyzed, the analyzing result is compared with the second screening threshold, and the optimal variable is screened out of the plural risk variables to serve as the target variable.

19. The device of any one of claims 1 to 18, wherein the methods employed to analyze the plural risk variables includes one or more of methods of linear regression, logistic regression and decision tree, wherein the logistic regression can employ the modes of forward regression, backward regression, and simultaneous forward regression and backward regression, wherein what mode is specifically employed can be decided by comparing plural results to see which mode is the best.

20. The device of any one of claims 1 to 19, wherein the plural risk variables are converted prior to being analyzed.

Date recue/ date received 2022-02-18

21. The device of any one of claims 1 to 20, wherein the coefficient of regression result of the plural risk variables is to setting the initial admission condition, wherein the product of the coefficient with ODDS of a corresponding variable is apparently lower than the average value, the value of a variable to which the value twice as much as the average value minus a standard difference corresponds can be set as a threshold, wherein each variable is performed with similar analysis before the corresponding threshold of each variable can be obtained.

22. The device of any one of claims 1 to 21, wherein the preset algorithm is employed to analyze the first user data, the first user data of a non-performing user is selected for analysis, wherein to analyze out features of the non-performing user to serve as reference to subsequent setting of the admission condition, wherein target users whose loans were once overdue, target users appearing on the blacklist of the core enterprise, and target users having no amount paid to picked up goods in six months are defined as inferior users or non-performing users, while target users clear of the above circumstances are defined as quality users.

23. The device of any one of claims 1 to 22, wherein the classified type to which each target user corresponds is judged as to whether the target user is a quality user or an inferior user, those target users that pertain to non-performing users are screened out, and the first user data to which these target users correspond is analyzed to obtain corresponding target variables.

24. The device of any one of claims 1 to 23, wherein the formula for calculating the credit line of the target user is:
credit line = max (basic line * coefficient ¨ Central Bank credit investigation used line, 0) Date recue/ date received 2022-02-18

25. The device of claim 24, wherein the Central Bank credit investigation used line indicates credit extension obtained in a name of the core enterprise, is the used credit line extracted from a credit investigation report of the Central Bank, wherein the basic line is the first line, and the Central Bank credit investigation used line is the second line, wherein the principle of setting credit lines of the target user is that the profitability of the user is capable of covering the existent credit line and the newly added credit line.

26. The device of any one of claims 24 to 25, wherein he basic line takes into consideration the safety of the line as provided, after the user has passed the target admission condition, the basic line is related to the repayment capability of the user, while the repayment capability is related to the income of the user.

27. The device of any one of claims 24 to 26, wherein the coefficient is determined after the basic line has been well determined, wherein the coefficient is a parameter with which to adequately readjust the basic line based on other information or newly appearing information after the basic line has been given, wherein higher the coefficient is, the better the credit of the core enterprise, wherein a lower coefficient indicates that the core enterprise falls short of perfect credit or leaves something to be desired of the credit.

28. The device of any one of claims 24 to 27, wherein after the coefficient has been determined, it is then considered to readjust the coefficient, wherein the readjustment of the coefficient mainly reflects information of the user relevant to recent change in credit risk, and information includes one or more of possible negative early warning signals, abrupt decrease in the sales, amount of money, and overdue circumstances of currently existing credits, wherein the readjustment of the coefficient aims to adequately lower the credit level of the client, control the credit line of the client, and even reject loan to the client when potential risk is encountered.

29. The device of any one of claims 24 to 28, wherein the main conditions relevant to readjustment of the coefficient comprises:
lowering the coefficient for one grade if there is an early warning signal against the target user;
Date recue/ date received 2022-02-18 lowering the coefficient for one grade if the goods picking-up amount of money of the target user decreases by 20% in the recent three months on a year-on-year basis;
lowering the coefficient for one grade if the goods picking-up amount of money of the target user decreases by 15% in last year on a year-on-year basis;
lowering the coefficient for one grade if abnormality occurs to the target user; and lowering the coefficient for one grade if the target user had overdue circumstances in the last two years, with the maximum overdue period being less than 15 days, and all overdue loans have been made good.

30. The device of any one of claims 1 to 29, wherein the first user data of the target user is requested from the first target platform, information of the downstream target users contents as the format of the required data and time duration of the required data can be ascertained.

31. A computer system for assessing user credit risk, the system comprising:
a memory, wherein the memory includes a nonvolatile storage medium, and an internal memory wherein the nonvolatile storage medium stores an operating system, a computer program and a database, and wherein the internal memory provides environment for the running of the operating system and the computer program in the nonvolatile storage medium;
a processor, wherein the processor is employed to provide computing and controlling capabilities;
a network interface connected to each other via a system bus and to connect to an external terminal via network for communication; and the computer program stored on the memory and operable on the processor, wherein the computer program realizes a method of optimizing an execution plan when it is executed by a processor.

32. The system of claim 31, wherein the processor executes the computer program comprising:

Date recue/ date received 2022-02-18 obtaining a first user data of a target user on a first target platform and a second user data of the target user on a second target platform;
employing a preset algorithm to analyze the first user data;
obtaining a target variable to which the target user corresponds, wherein the target variable includes a variable describing credit risk of the target user;
determining an initial admission condition of credit according to the target variable and a preset threshold;
readjusting the initial admission condition according to the second user data;

obtaining a target admission condition; and determining a credit line of each target user that conforms to the target admission condition according to the first user data and the second user data.

33. The system of claim 32, wherein employing the preset algorithm to analyze the first user data, and obtaining the target variable to which the target user corresponds comprises:
constructing a risk variable according to the first user data and a preset rule, wherein the risk variable is used to describe credit risk of the target user;
employing the preset algorithm to analyze the risk variable according to preset dimensions; and screening the target variable out of the risk variable according to an analyzing result and a preset screening threshold.

34. The system of claim 33, wherein the preset algorithm comprises a first preset algorithm and a second preset algorithm, wherein the preset screening threshold comprises a preset first screening threshold and a preset second screening threshold comprises:

Date recue/ date received 2022-02-18 employing the preset algorithm to analyze the risk variable according to the preset dimensions;
screening the target variable out of the risk variable according to the analyzing result and the preset screening threshold comprises:
employing the first preset algorithm to analyze the risk variable according to the preset dimensions, and obtaining a first analyzing result;
screening out a plurality of risk variables according to the first analyzing result and the preset first screening threshold;
employing the second preset algorithm to analyze the plural risk variables;
obtaining a second analyzing result; and screening the target variable out of the plural risk variables according to the second analyzing result and the preset second screening threshold.

35. The system of claim 34, wherein the second preset algorithm comprises a regression algorithm, wherein employing the second preset algorithm to analyze the plural risk variables, and obtaining the second analyzing result comprises:
performing a discretization process on the plural risk variables;
obtaining fragments to which the plural risk variables correspond;
employing the regression algorithm to perform regression analysis on the fragments; and obtaining the second analyzing result.

36. The system of any one of claims 31 to 35, wherein employing the preset algorithm to analyze the first user data, and obtaining the target variable to which the target user corresponds comprises:
classifying the target user according to a preset classifying rule;

Date recue/ date received 2022-02-18 obtaining a classified type to which each target user corresponds;
determining any target user that satisfies a preset analyzing condition according to the classified type;
employing a preset algorithm to analyze the first user data wherein the target user that satisfies the analyzing condition corresponds; and obtaining a corresponding target variable.

37. The system of any one of claims 31 to 36, wherein the step of determining the credit line of each target user that conforms to the target admission condition according to the first user data and the second user data comprises:
determining a first line of each target user that conforms to the target admission condition and a coefficient to which the first line corresponds according to the first user data and a preset line table;
determining a second line of each target user that conforms to the target admission condition according to the second user data; and calculating to obtain the credit line of each target user that conforms to the target admission condition according to the first line, the coefficient, and the second line.

38. The system of any one of claims 31 to 37, wherein obtaining the first user data of the target user on the first target platform comprises:
sending a data request requesting the first user data of the target user to the first target platform, wherein the data request includes a data format of the first user data; and receiving the first user data that conforms to the data format returned by the first target platform.

Date recue/ date received 2022-02-18

39. The system of any one of claims 31 to 38, wherein the target user includes any one or more of a small and micro enterprise, wherein the first target platform includes a core enterprise in a supply chain, wherein the core enterprise in the supply chain, includes any one or more of times and quota of picking-up per month, seasonal fluctuation rates of goods, categories of goods, and cooperation durations with the core enterprise, wherein the first user data includes target user data possessed by the core enterprise, wherein target user data possessed by the core enterprise includes any one or more of the locations where target users businesses concentrate, and the circumstances of downstream target users, wherein the second target platform includes a credit investigation company, and wherein the second user data includes credit investigation data of the target user.

40. The system of any one of claims 31 to 39, wherein the first user data provided by the first target platform has been obtained, based on the first user data to analyze the circumstances of the target user, wherein the preset algorithm can be employed to analyze the first user data, and the analyzed contents includes any one or more of distribution of various features of the target user, distribution of circumstances of the target user whose credits have been overdue in other banks, distribution of rewards and punishments for delivery, and distribution of delivery amounts of the target user, wherein the target variable is reasonably set in conjunction with the distribution circumstances of data and overdue features.

41. The system of any one of claims 31 to 40, wherein the target variable has been determined, the initial admission condition is set in conjunction with the target variable and the preset threshold, wherein setup of the initial admission condition here is combined with the analyzing result.
Date recue/ date received 2022-02-18

42. The system of any one of claims 31 to 41, wherein the initial admission condition includes any one or more of no appearance on a blacklist of the core enterprise of the supply chain, exclusion of target users with inferior or severe overdue circumstance, over 7 days, on a credit list of the core enterprise of the supply chain, exclusion of provinces or cities to which a business is temporarily not extended, due to reasons of small business quantity, undeveloped economy or inferior credit environment, exclusion of target users whose yearly picking-ups of goods are less than three months, exclusion of target users who pick up goods in just one month or less than one month in the recent three months, exclusion of target users whose amount of goods picked up is low, exclusion of target users whose order completion rate is lower than 30%, exclusion of target users whose amount of goods picked up is reduced over 50% on a year-on-year or month-on-month basis within one year, and exclusion of target users who opened accounts at the core enterprise less than one year.

43. The system of any one of claims 31 to 42, wherein third-party credit investigation company and a central bank provides information relevant to the downstream target user, including one or more of business information of the target user, judicially involved information, social loan information, wherein the information provided by the core enterprise and the credit investigation information are combined to readjust the initial admission condition obtained to obtain target admission data.

44. The system of any one of claims 31 to 43, wherein by analyzing the first user data obtained from the first target platform in combination with experiences acquired from other projects can be sorted out a upper limit of creditable line of each target user, wherein from information provided by the Central Bank or the third-party credit investigation company, the second user data of the second target platform, can be analyzed out credit circumstances of each target user in other banks, including one or more of creditable line, already used line, overdue or collection information, wherein the actual creditable line of each target user can be obtained by combining information of these two aspects.

Date recue/ date received 2022-02-18

45. The system of any one of claims 31 to 44, wherein constructing the risk variable is to construct those potential variables that are capable of reflecting credit levels of the target user, wherein the risk variable constructed according to the first user data and the preset rule includes one or more of number of months within a year in which orders are completed, goods picking-up quota in one year, credit balance, number of months in which goods are picked up in the last year, number of months in which goods are picked up in the last six months, number of months in which goods are picked up in last three months, goods picking-up quota in last year, goods picking-up quota in last six months, goods picking-up quota in last three months, year-on-year growth rate of goods picked up in last year, month-on-month growth rate of goods picked up in the last year, year-on-year growth rate of goods picked up in the last six months, month-on-month growth rate of goods picked up in the last six months, year-on-year growth rate of goods picked up in the last three months, month-on-month growth rate of goods picked up in the last three months, region, province, city where the target user resides, time when the target user opened account, the amount of orders completed in the last year.

46. The system of any one of claims 31 to 45, wherein the preset algorithm is employed to analyze the risk variable, dimension information to be analyzed can be set in advance according to actual requirements, wherein the analyzing result as obtained is compared with the preset screening threshold, and wherein the target variable that conforms to the requirements is screened out of the risk variable.

47. The system of any one of claims 31 to 46, wherein employing the preset algorithm to analyze the risk variable according to preset dimensions includes analyzing a single risk variable and analyzing a plurality of risk variables, and employing the first preset algorithm to analyze the risk variable according to preset dimensions comprising:
analyzing the number of months within one year in which the target user completes orders, distribution of months, and relation between goods picking-up quota in a total year and non-performance of the target user;

Date recue/ date received 2022-02-18 analyzing credit balance distribution, overdue or non-performing balance distribution of the target user, and relation between credit balance and overdue or non-performance on the basis of the first user data provided by the first target platform, wherein severely overdue and non-perfonning users are avoided;
analyzing number of months in the last year in which goods are picked up, number of months in the last six months in which goods are picked up, and number of months in the last three months in which goods are picked up by each target user, and analyzing relation between these data and change in goods picking-up quota of the target user and possible overdue circumstances;
analyzing relation between non-performance and year-on-year and month-on-month growth rates of goods picked up in the last year, the year-on-year and month-on-month growth rates of goods picked up in the last six months, and the year-on-year and month-on-month growth rates of goods picked up in the last three months and analyzing the relation between the growth rates and change in goods picking-up quota of the target user in a total year and possible overdue circumstances;
analyzing the relation between the region where target user resides and the non-performance of the target user, wherein region selection include one or more of potential business volumes, credit environments and economic development levels of the region;
analyzing the relation between time when the target user opened account and the non-performance of target user, wherein target user's account is opened at least one year before sufficient goods picking-up history of the target user can be accumulated for analysis;
analyzing relation between the non-performance of the target user and the goods picking-up quota in the last year, the goods picking-up quota in the last sixth months, and the goods picking-up quota in the last three months of the target user; and analyzing the relation between the order completion rate of the target user in the core enterprise and the non-performance of the target user.

Date recue/ date received 2022-02-18

48. The system of any one of claims 31 to 47, wherein the relations analyzed are reflected through statistics, including statistic indices AUC and KS, wherein the first screening threshold is provided to the statistic index of each risk variable, and the risk variables exceeding the first screening threshold are preliminarily selected to pass, wherein passing the first round of screening, plural risk variables possessing capability to apparently differentiating target user risks are screened out to serve as basis for analyzing the plurality of risk variables.

49. The system of any one of claims 31 to 48, wherein analyzing the plurality of risk variables is to find out a collection of variables most capable of differentiating target user credit levels, wherein the second screening threshold can be preset while the plural risk variables are being analyzed, the analyzing result is compared with the second screening threshold, and the optimal variable is screened out of the plural risk variables to serve as the target variable.

50. The system of any one of claims 31 to 49, wherein the methods employed to analyze the plural risk variables includes one or more of methods of linear regression, logistic regression and decision tree, wherein the logistic regression can employ the modes of forward regression, backward regression, and simultaneous forward regression and backward regression, wherein what mode is specifically employed can be decided by comparing plural results to see which mode is the best.

51. The system of any one of claims 31 to 50, wherein the plural risk variables are converted prior to being analyzed.

52. The system of any one of claims 31 to 51, wherein the coefficient of regression result of the plural risk variables is to setting the initial admission condition, wherein the product of the coefficient with ODDS of a corresponding variable is apparently lower than the average value, the value of a variable to which the value twice as much as the average value minus a standard difference corresponds can be set as a threshold, wherein each variable is performed with similar analysis before the corresponding threshold of each variable can be obtained.

Date recue/ date received 2022-02-18

53. The system of any one of claims 31 to 52, wherein the preset algorithm is employed to analyze the first user data, the first user data of a non-performing user is selected for analysis, wherein to analyze out features of the non-performing user to serve as reference to subsequent setting of the admission condition, wherein target users whose loans were once overdue, target users appearing on the blacklist of the core enterprise, and target users having no amount paid to picked up goods in six months are defined as inferior users or non-performing users, while target users clear of the above circumstances are defined as quality users.

54. The system of any one of claims 31 to 53, wherein the classified type to which each target user corresponds is judged as to whether the target user is a quality user or an inferior user, those target users that pertain to non-performing users are screened out, and the first user data to which these target users correspond is analyzed to obtain corresponding target variables.

55. The system of any one of claims 31 to 54, wherein the formula for calculating the credit line of the target user is:
credit line = max (basic line * coefficient ¨ Central Bank credit investigation used line, 0)

56. The system of claim 55, wherein the Central Bank credit investigation used line indicates credit extension obtained in a name of the core enterprise, is the used credit line extracted from a credit investigation report of the Central Bank, wherein the basic line is the first line, and the Central Bank credit investigation used line is the second line, wherein the principle of setting credit lines of the target user is that the profitability of the user is capable of covering the existent credit line and the newly added credit line.

57. The system of any one of claims 55 to 56, wherein he basic line takes into consideration the safety of the line as provided, after the user has passed the target admission condition, the basic line is related to the repayment capability of the user, while the repayment capability is related to the income of the user.
Date recue/ date received 2022-02-18

58. The system of any one of claims 55 to 57, wherein the coefficient is determined after the basic line has been well determined, wherein the coefficient is a parameter with which to adequately readjust the basic line based on other information or newly appearing information after the basic line has been given, wherein higher the coefficient is, the better the credit of the core enterprise, wherein a lower coefficient indicates that the core enterprise falls short of perfect credit or leaves something to be desired of the credit.

59. The system of any one of claims 55 to 56, wherein after the coefficient has been determined, it is then considered to readjust the coefficient, wherein the readjustment of the coefficient mainly reflects information of the user relevant to recent change in credit risk, and information includes one or more of possible negative early warning signals, abrupt decrease in the sales, amount of money, and overdue circumstances of currently existing credits, wherein the readjustment of the coefficient aims to adequately lower the credit level of the client, control the credit line of the client, and even reject loan to the client when potential risk is encountered.

60. The system of any one of claims 55 to 59, wherein the main conditions relevant to readjustment of the coefficient comprises:
lowering the coefficient for one grade if there is an early warning signal against the target user;
lowering the coefficient for one grade if the goods picking-up amount of money of the target user decreases by 20% in the recent three months on a year-on-year basis;
lowering the coefficient for one grade if the goods picking-up amount of money of the target user decreases by 15% in last year on a year-on-year basis;
lowering the coefficient for one grade if abnormality occurs to the target user; and lowering the coefficient for one grade if the target user had overdue circumstances in the last two years, with the maximum overdue period being less than 15 days, and all overdue loans have been made good.

Date recue/ date received 2022-02-18

61. The system of any one of claims 31 to 60, wherein the first user data of the target user is requested from the first target platform, information of the downstream target users contents as the format of the required data and time duration of the required data can be ascertained.

62. A method for assessing user credit risk, the method comprising:
obtaining a first user data of a target user on a first target platform and a second user data of the target user on a second target platform;
employing a preset algorithm to analyze the first user data;
obtaining a target variable to which the target user corresponds, wherein the target variable includes a variable describing credit risk of the target user;
determining an initial admission condition of credit according to the target variable and a preset threshold;
readjusting the initial admission condition according to the second user data;

obtaining a target admission condition; and determining a credit line of each target user that conforms to the target admission condition according to the first user data and the second user data.

63. The method of claim 62, wherein employing the preset algorithm to analyze the first user data, and obtaining the target variable to which the target user corresponds comprises:
constructing a risk variable according to the first user data and a preset rule, wherein the risk variable is used to describe credit risk of the target user;
employing the preset algorithm to analyze the risk variable according to preset dimensions; and screening the target variable out of the risk variable according to an analyzing result and a preset screening threshold.

Date recue/ date received 2022-02-18

64. The method of claim 63, wherein the preset algorithm comprises a first preset algorithm and a second preset algorithm, wherein the preset screening threshold comprises a preset first screening threshold and a preset second screening threshold comprises:
employing the preset algorithm to analyze the risk variable according to the preset dimensions;
screening the target variable out of the risk variable according to the analyzing result and the preset screening threshold comprises:
employing the first preset algorithm to analyze the risk variable according to the preset dimensions, and obtaining a first analyzing result;
screening out a plurality of risk variables according to the first analyzing result and the preset first screening threshold;
employing the second preset algorithm to analyze the plural risk variables;
obtaining a second analyzing result; and screening the target variable out of the plural risk variables according to the second analyzing result and the preset second screening threshold.

65. The method of claim 64, wherein the second preset algorithm comprises a regression algorithm, wherein employing the second preset algorithm to analyze the plural risk variables, and obtaining the second analyzing result comprises:
performing a discretization process on the plural risk variables;
obtaining fragments to which the plural risk variables correspond;
employing the regression algorithm to perform regression analysis on the fragments; and obtaining the second analyzing result.

Date recue/ date received 2022-02-18

66. The method of any one of claims 62 to 65, wherein employing the preset algorithm to analyze the first user data, and obtaining the target variable to which the target user corresponds comprises:
classifying the target user according to a preset classifying rule;
obtaining a classified type to which each target user corresponds;
determining any target user that satisfies a preset analyzing condition according to the classified type;
employing a preset algorithm to analyze the first user data wherein the target user that satisfies the analyzing condition corresponds; and obtaining a corresponding target variable.

67. The method of any one of claims 62 to 65, wherein the step of determining the credit line of each target user that conforms to the target admission condition according to the first user data and the second user data comprises:
determining a first line of each target user that conforms to the target admission condition and a coefficient to which the first line corresponds according to the first user data and a preset line table;
determining a second line of each target user that conforms to the target admission condition according to the second user data; and calculating to obtain the credit line of each target user that conforms to the target admission condition according to the first line, the coefficient, and the second line.

68. The method of any one of claims 62 to 65, wherein obtaining the first user data of the target user on the first target platform comprises:
sending a data request requesting the first user data of the target user to the first target platform, wherein the data request includes a data format of the first user data; and Date recue/ date received 2022-02-18 receiving the first user data that conforms to the data format returned by the first target platform.

69. The method of any one of claims 62 to 68, wherein the target user includes any one or more of a small and micro enterprise, wherein the first target platform includes a core enterprise in a supply chain, wherein the core enterprise in the supply chain, includes any one or more of times and quota of picking-up per month, seasonal fluctuation rates of goods, categories of goods, and cooperation durations with the core enterprise, wherein the first user data includes target user data possessed by the core enterprise, wherein target user data possessed by the core enterprise includes any one or more of the locations where target users businesses concentrate, and the circumstances of downstream target users, wherein the second target platform includes a credit investigation company, and wherein the second user data includes credit investigation data of the target user.

70. The method of any one of claims 62 to 69, wherein the first user data provided by the first target platform has been obtained, based on the first user data to analyze the circumstances of the target user, wherein the preset algorithm can be employed to analyze the first user data, and the analyzed contents includes any one or more of distribution of various features of the target user, distribution of circumstances of the target user whose credits have been overdue in other banks, distribution of rewards and punishments for delivery, and distribution of delivery amounts of the target user, wherein the target variable is reasonably set in conjunction with the distribution circumstances of data and overdue features.

71. The method of any one of claims 62 to 70, wherein the target variable has been determined, the initial admission condition is set in conjunction with the target variable and the preset threshold, wherein setup of the initial admission condition here is combined with the analyzing result.
Date recue/ date received 2022-02-18

72. The method of any one of claims 62 to 71, wherein the initial admission condition includes any one or more of no appearance on a blacklist of the core enterprise of the supply chain, exclusion of target users with inferior or severe overdue circumstance, over 7 days, on a credit list of the core enterprise of the supply chain, exclusion of provinces or cities to which a business is temporarily not extended, due to reasons of small business quantity, undeveloped economy or inferior credit environment, exclusion of target users whose yearly picking-ups of goods are less than three months, exclusion of target users who pick up goods in just one month or less than one month in the recent three months, exclusion of target users whose amount of goods picked up is low, exclusion of target users whose order completion rate is lower than 30%, exclusion of target users whose amount of goods picked up is reduced over 50% on a year-on-year or month-on-month basis within one year, and exclusion of target users who opened accounts at the core enterprise less than one year.

73. The method of any one of claims 62 to 72, wherein third-party credit investigation company and a central bank provides information relevant to the downstream target user, including one or more of business information of the target user, judicially involved information, social loan information, wherein the information provided by the core enterprise and the credit investigation information are combined to readjust the initial admission condition obtained to obtain target admission data.

74. The method of any one of claims 62 to 73, wherein by analyzing the first user data obtained from the first target platform in combination with experiences acquired from other projects can be sorted out a upper limit of creditable line of each target user, wherein from information provided by the Central Bank or the third-party credit investigation company, the second user data of the second target platform, can be analyzed out credit circumstances of each target user in other banks, including one or more of creditable line, already used line, overdue or collection information, wherein the actual creditable line of each target user can be obtained by combining information of these two aspects.

Date recue/ date received 2022-02-18

75. The method of any one of claims 62 to 74, wherein constructing the risk variable is to construct those potential variables that are capable of reflecting credit levels of the target user, wherein the risk variable constructed according to the first user data and the preset rule includes one or more of number of months within a year in which orders are completed, goods picking-up quota in one year, credit balance, number of months in which goods are picked up in the last year, number of months in which goods are picked up in the last six months, number of months in which goods are picked up in last three months, goods picking-up quota in last year, goods picking-up quota in last six months, goods picking-up quota in last three months, year-on-year growth rate of goods picked up in last year, month-on-month growth rate of goods picked up in the last year, year-on-year growth rate of goods picked up in the last six months, month-on-month growth rate of goods picked up in the last six months, year-on-year growth rate of goods picked up in the last three months, month-on-month growth rate of goods picked up in the last three months, region, province, city where the target user resides, time when the target user opened account, the amount of orders completed in the last year.

76. The method of any one of claims 62 to 75, wherein the preset algorithm is employed to analyze the risk variable, dimension information to be analyzed can be set in advance according to actual requirements, wherein the analyzing result as obtained is compared with the preset screening threshold, and wherein the target variable that conforms to the requirements is screened out of the risk variable.

77. The method of any one of claims 62 to 76, wherein employing the preset algorithm to analyze the risk variable according to preset dimensions includes analyzing a single risk variable and analyzing a plurality of risk variables, and employing the first preset algorithm to analyze the risk variable according to preset dimensions comprising:
analyzing the number of months within one year in which the target user completes orders, distribution of months, and relation between goods picking-up quota in a total year and non-performance of the target user;

Date recue/ date received 2022-02-18 analyzing credit balance distribution, overdue or non-performing balance distribution of the target user, and relation between credit balance and overdue or non-performance on the basis of the first user data provided by the first target platform, wherein severely overdue and non-perfomiing users are avoided;
analyzing number of months in the last year in which goods are picked up, number of months in the last six months in which goods are picked up, and number of months in the last three months in which goods are picked up by each target user, and analyzing relation between these data and change in goods picking-up quota of the target user and possible overdue circumstances;
analyzing relation between non-performance and year-on-year and month-on-month growth rates of goods picked up in the last year, the year-on-year and month-on-month growth rates of goods picked up in the last six months, and the year-on-year and month-on-month growth rates of goods picked up in the last three months and analyzing the relation between the growth rates and change in goods picking-up quota of the target user in a total year and possible overdue circumstances;
analyzing the relation between the region where target user resides and the non-performance of the target user, wherein region selection include one or more of potential business volumes, credit environments and economic development levels of the region;
analyzing the relation between time when the target user opened account and the non-performance of target user, wherein target user's account is opened at least one year before sufficient goods picking-up history of the target user can be accumulated for analysis;
analyzing relation between the non-performance of the target user and the goods picking-up quota in the last year, the goods picking-up quota in the last sixth months, and the goods picking-up quota in the last three months of the target user; and Date recue/ date received 2022-02-18 analyzing the relation between the order completion rate of the target user in the core enterprise and the non-performance of the target user.

78. The method of any one of claims 62 to 77, wherein the relations analyzed are reflected through statistics, including statistic indices AUC and KS, wherein the first screening threshold is provided to the statistic index of each risk variable, and the risk variables exceeding the first screening threshold are preliminarily selected to pass, wherein passing the first round of screening, plural risk variables possessing capability to apparently differentiating target user risks are screened out to serve as basis for analyzing the plurality of risk variables.

79. The method of any one of claims 62 to 78, wherein analyzing the plurality of risk variables is to find out a collection of variables most capable of differentiating target user credit levels, wherein the second screening threshold can be preset while the plural risk variables are being analyzed, the analyzing result is compared with the second screening threshold, and the optimal variable is screened out of the plural risk variables to serve as the target variable.

80. The method of any one of claims 62 to 74, wherein the methods employed to analyze the plural risk variables includes one or more of methods of linear regression, logistic regression and decision tree, wherein the logistic regression can employ the modes of forward regression, backward regression, and simultaneous forward regression and backward regression, wherein what mode is specifically employed can be decided by comparing plural results to see which mode is the best.

81. The method of any one of claims 62 to 80, wherein the plural risk variables are converted prior to being analyzed.

Date recue/ date received 2022-02-18

82. The method of any one of claims 62 to 81, wherein the coefficient of regression result of the plural risk variables is to setting the initial admission condition, wherein the product of the coefficient with ODDS of a corresponding variable is apparently lower than the average value, the value of a variable to which the value twice as much as the average value minus a standard difference corresponds can be set as a threshold, wherein each variable is performed with similar analysis before the corresponding threshold of each variable can be obtained.

83. The method of any one of claims 62 to 82, wherein the preset algorithm is employed to analyze the first user data, the first user data of a non-performing user is selected for analysis, wherein to analyze out features of the non-performing user to serve as reference to subsequent setting of the admission condition, wherein target users whose loans were once overdue, target users appearing on the blacklist of the core enterprise, and target users having no amount paid to picked up goods in six months are defined as inferior users or non-performing users, while target users clear of the above circumstances are defined as quality users.

84. The method of any one of claims 62 to 83, wherein the classified type to which each target user corresponds is judged as to whether the target user is a quality user or an inferior user, those target users that pertain to non-performing users are screened out, and the first user data to which these target users correspond is analyzed to obtain corresponding target variables.

85. The method of any one of claims 62 to 84, wherein the formula for calculating the credit line of the target user is:
credit line = max (basic line * coefficient ¨ Central Bank credit investigation used line, 0) Date recue/ date received 2022-02-18

86. The method of claim 85, wherein the Central Bank credit investigation used line indicates credit extension obtained in a name of the core enterprise, is the used credit line extracted from a credit investigation report of the Central Bank, wherein the basic line is the first line, and the Central Bank credit investigation used line is the second line, wherein the principle of setting credit lines of the target user is that the profitability of the user is capable of covering the existent credit line and the newly added credit line.

87. The method of any one of claims 85 to 86, wherein he basic line takes into consideration the safety of the line as provided, after the user has passed the target admission condition, the basic line is related to the repayment capability of the user, while the repayment capability is related to the income of the user.

88. The method of any one of claims 85 to 87, wherein the coefficient is determined after the basic line has been well determined, wherein the coefficient is a parameter with which to adequately readjust the basic line based on other information or newly appearing information after the basic line has been given, wherein higher the coefficient is, the better the credit of the core enterprise, wherein a lower coefficient indicates that the core enterprise falls short of perfect credit or leaves something to be desired of the credit.

89. The method of any one of claims 85 to 88, wherein after the coefficient has been determined, it is then considered to readjust the coefficient, wherein the readjustment of the coefficient mainly reflects information of the user relevant to recent change in credit risk, and information includes one or more of possible negative early warning signals, abrupt decrease in the sales, amount of money, and overdue circumstances of currently existing credits, wherein the readjustment of the coefficient aims to adequately lower the credit level of the client, control the credit line of the client, and even reject loan to the client when potential risk is encountered.

90. The method of any one of claims 85 to 89, wherein the main conditions relevant to readjustment of the coefficient comprises:
lowering the coefficient for one grade if there is an early warning signal against the target user;

Date recue/ date received 2022-02-18 lowering the coefficient for one grade if the goods picking-up amount of money of the target user decreases by 20% in the recent three months on a year-on-year basis;
lowering the coefficient for one grade if the goods picking-up amount of money of the target user decreases by 15% in last year on a year-on-year basis;
lowering the coefficient for one grade if abnormality occurs to the target user; and lowering the coefficient for one grade if the target user had overdue circumstances in the last two years, with the maximum overdue period being less than 15 days, and all overdue loans have been made good.

91. The method of any one of claims 62 to 90, wherein the first user data of the target user is requested from the first target platform, information of the downstream target users contents as the format of the required data and time duration of the required data can be ascertained.

92. A computer-readable storage medium storing a computer program thereon configured to:
obtain a first user data of a target user on a first target platform and a second user data of the target user on a second target platform;
employ a preset algorithm to analyze the first user data;
obtain a target variable to which the target user corresponds, wherein the target variable includes a variable describing credit risk of the target user;
determine an initial admission condition of credit according to the target variable and a preset threshold;
readjust the initial admission condition according to the second user data;
obtain a target admission condition; and determine a credit line of each target user that conforms to the target admission condition according to the first user data and the second user data.

Date recue/ date received 2022-02-18

93. The storage medium of claim 92, wherein employing the preset algorithm to analyze the first user data, and obtaining the target variable to which the target user corresponds comprises:
constructing a risk variable according to the first user data and a preset rule, wherein the risk variable is used to describe credit risk of the target user;
employing the preset algorithm to analyze the risk variable according to preset dimensions; and screening the target variable out of the risk variable according to an analyzing result and a preset screening threshold.

94. The storage medium of claim 93, wherein the preset algorithm comprises a first preset algorithm and a second preset algorithm, wherein the preset screening threshold comprises a preset first screening threshold and a preset second screening threshold comprises:
employing the preset algorithm to analyze the risk variable according to the preset dimensions;
screening the target variable out of the risk variable according to the analyzing result and the preset screening threshold comprises:
employing the first preset algorithm to analyze the risk variable according to the preset dimensions, and obtaining a first analyzing result;
screening out a plurality of risk variables according to the first analyzing result and the preset first screening threshold;
employing the second preset algorithm to analyze the plural risk variables;
obtaining a second analyzing result; and screening the target variable out of the plural risk variables according to the second analyzing result and the preset second screening threshold.

Date recue/ date received 2022-02-18

95. The storage medium of claim 94, wherein the second preset algorithm comprises a regression algorithm, wherein employing the second preset algorithm to analyze the plural risk variables, and obtaining the second analyzing result comprises:
performing a discretization process on the plural risk variables;
obtaining fragments to which the plural risk variables correspond;
employing the regression algorithm to perform regression analysis on the fragments; and obtaining the second analyzing result.

96. The storage medium of any one of claims 92 to 95, wherein employing the preset algorithm to analyze the first user data, and obtaining the target variable to which the target user corresponds comprises:
classifying the target user according to a preset classifying rule;
obtaining a classified type to which each target user corresponds;
determining any target user that satisfies a preset analyzing condition according to the classified type;
employing a preset algorithm to analyze the first user data wherein the target user that satisfies the analyzing condition corresponds; and obtaining a corresponding target variable.

97. The storage medium of any one of claims 92 to 96, wherein the step of determining the credit line of each target user that conforms to the target admission condition according to the first user data and the second user data comprises:
determining a first line of each target user that conforms to the target admission condition and a coefficient to which the first line corresponds according to the first user data and a preset line table;

Date recue/ date received 2022-02-18 determining a second line of each target user that conforms to the target admission condition according to the second user data; and calculating to obtain the credit line of each target user that conforms to the target admission condition according to the first line, the coefficient, and the second line.

98. The storage medium of any one of claims 92 to 97, wherein obtaining the first user data of the target user on the first target platform comprises:
sending a data request requesting the first user data of the target user to the first target platform, wherein the data request includes a data format of the first user data; and receiving the first user data that conforms to the data format returned by the first target platform.

99. The storage medium of any one of claims 92 to 98, wherein the target user includes any one or more of a small and micro enterprise, wherein the first target platform includes a core enterprise in a supply chain, wherein the core enterprise in the supply chain, includes any one or more of times and quota of picking-up per month, seasonal fluctuation rates of goods, categories of goods, and cooperation durations with the core enterprise, wherein the first user data includes target user data possessed by the core enterprise, wherein target user data possessed by the core enterprise includes any one or more of the locations where target users businesses concentrate, and the circumstances of downstream target users, wherein the second target platform includes a credit investigation company, and wherein the second user data includes credit investigation data of the target user.
Date recue/ date received 2022-02-18

100. The storage medium of any one of claims 92 to 99, wherein the first user data provided by the first target platform has been obtained, based on the first user data to analyze the circumstances of the target user, wherein the preset algorithm can be employed to analyze the first user data, and the analyzed contents includes any one or more of distribution of various features of the target user, distribution of circumstances of the target user whose credits have been overdue in other banks, distribution of rewards and punishments for delivery, and distribution of delivery amounts of the target user, wherein the target variable is reasonably set in conjunction with the distribution circumstances of data and overdue features.

101. The storage medium of any one of claims 92 to 100, wherein the target variable has been determined, the initial admission condition is set in conjunction with the target variable and the preset threshold, wherein setup of the initial admission condition here is combined with the analyzing result.

102. The storage medium of any one of claims 92 to 101, wherein the initial admission condition includes any one or more of no appearance on a blacklist of the core enterprise of the supply chain, exclusion of target users with inferior or severe overdue circumstance, over 7 days, on a credit list of the core enterprise of the supply chain, exclusion of provinces or cities to which a business is temporarily not extended, due to reasons of small business quantity, undeveloped economy or inferior credit environment, exclusion of target users whose yearly picking-ups of goods are less than three months, exclusion of target users who pick up goods in just one month or less than one month in the recent three months, exclusion of target users whose amount of goods picked up is low, exclusion of target users whose order completion rate is lower than 30%, exclusion of target users whose amount of goods picked up is reduced over 50% on a year-on-year or month-on-month basis within one year, and exclusion of target users who opened accounts at the core enterprise less than one year.

Date recue/ date received 2022-02-18

103. The storage medium of any one of claims 92 to 102, wherein third-party credit investigation company and a central bank provides information relevant to the downstream target user, including one or more of business information of the target user, judicially involved information, social loan information, wherein the information provided by the core enterprise and the credit investigation information are combined to readjust the initial admission condition obtained to obtain target admission data.

104. The storage medium of any one of claims 92 to 103, wherein by analyzing the first user data obtained from the first target platform in combination with experiences acquired from other projects can be sorted out a upper limit of creditable line of each target user, wherein from information provided by the Central Bank or the third-party credit investigation company, the second user data of the second target platform, can be analyzed out credit circumstances of each target user in other banks, including one or more of creditable line, already used line, overdue or collection information, wherein the actual creditable line of each target user can be obtained by combining information of these two aspects.

105. The storage medium of any one of claims 92 to 104, wherein constructing the risk variable is to construct those potential variables that are capable of reflecting credit levels of the target user, wherein the risk variable constructed according to the first user data and the preset rule includes one or more of number of months within a year in which orders are completed, goods picking-up quota in one year, credit balance, number of months in which goods are picked up in the last year, number of months in which goods are picked up in the last six months, number of months in which goods are picked up in last three months, goods picking-up quota in last year, goods picking-up quota in last six months, goods picking-up quota in last three months, year-on-year growth rate of goods picked up in last year, month-on-month growth rate of goods picked up in the last year, year-on-year growth rate of goods picked up in the last six months, month-on-month growth rate of goods picked up in the last six months, year-on-year growth rate of goods picked up in the last three months, month-on-month growth rate of goods picked up in the last three months, region, province, city where the target user resides, time when the target user opened account, the amount of orders completed in the last year.

Date recue/ date received 2022-02-18

106. The storage medium of any one of claims 92 to 105, wherein the preset algorithm is employed to analyze the risk variable, dimension information to be analyzed can be set in advance according to actual requirements, wherein the analyzing result as obtained is compared with the preset screening threshold, and wherein the target variable that conforms to the requirements is screened out of the risk variable.

107. The storage medium of any one of claims 92 to 106, wherein employing the preset algorithm to analyze the risk variable according to preset dimensions includes analyzing a single risk variable and analyzing a plurality of risk variables, and employing the first preset algorithm to analyze the risk variable according to preset dimensions comprising:
analyzing the number of months within one year in which the target user completes orders, distribution of months, and relation between goods picking-up quota in a total year and non-performance of the target user;
analyzing credit balance distribution, overdue or non-performing balance distribution of the target user, and relation between credit balance and overdue or non-performance on the basis of the first user data provided by the first target platform, wherein severely overdue and non-performing users are avoided;
analyzing number of months in the last year in which goods are picked up, number of months in the last six months in which goods are picked up, and number of months in the last three months in which goods are picked up by each target user, and analyzing relation between these data and change in goods picking-up quota of the target user and possible overdue circumstances;
analyzing relation between non-performance and year-on-year and month-on-month growth rates of goods picked up in the last year, the year-on-year and month-on-month growth rates of goods picked up in the last six months, and the year-on-year and month-on-month growth rates of goods picked up in the last three months and analyzing the relation between the growth rates and change in goods picking-up quota of the target user in a total year and possible overdue circumstances;

Date recue/ date received 2022-02-18 analyzing the relation between the region where target user resides and the non-performance of the target user, wherein region selection include one or more of potential business volumes, credit environments and economic development levels of the region;
analyzing the relation between time when the target user opened account and the non-performance of target user, wherein target user's account is opened at least one year before sufficient goods picking-up history of the target user can be accumulated for analysis;
analyzing relation between the non-performance of the target user and the goods picking-up quota in the last year, the goods picking-up quota in the last sixth months, and the goods picking-up quota in the last three months of the target user; and analyzing the relation between the order completion rate of the target user in the core enterprise and the non-performance of the target user.

108. The storage medium of any one of claims 92 to 107, wherein the relations analyzed are reflected through statistics, including statistic indices AUC and KS, wherein the first screening threshold is provided to the statistic index of each risk variable, and the risk variables exceeding the first screening threshold are preliminarily selected to pass, wherein passing the first round of screening, plural risk variables possessing capability to apparently differentiating target user risks are screened out to serve as basis for analyzing the plurality of risk variables.

109. The storage medium of any one of claims 92 to 108, wherein analyzing the plurality of risk variables is to find out a collection of variables most capable of differentiating target user credit levels, wherein the second screening threshold can be preset while the plural risk variables are being analyzed, the analyzing result is compared with the second screening threshold, and the optimal variable is screened out of the plural risk variables to serve as the target variable.

Date recue/ date received 2022-02-18

110. The storage medium of any one of claims 92 to 109, wherein the methods employed to analyze the plural risk variables includes one or more of methods of linear regression, logistic regression and decision tree, wherein the logistic regression can employ the modes of forward regression, backward regression, and simultaneous forward regression and backward regression, wherein what mode is specifically employed can be decided by comparing plural results to see which mode is the best.

111. The storage medium of any one of claims 92 to 110, wherein the plural risk variables are converted prior to being analyzed.

112. The storage medium of any one of claims 92 to 111, wherein the coefficient of regression result of the plural risk variables is to setting the initial admission condition, wherein the product of the coefficient with ODDS of a corresponding variable is apparently lower than the average value, the value of a variable to which the value twice as much as the average value minus a standard difference corresponds can be set as a threshold, wherein each variable is performed with similar analysis before the corresponding threshold of each variable can be obtained.

113. The storage medium of any one of claims 92 to 112, wherein the preset algorithm is employed to analyze the first user data, the first user data of a non-performing user is selected for analysis, wherein to analyze out features of the non-performing user to serve as reference to subsequent setting of the admission condition, wherein target users whose loans were once overdue, target users appearing on the blacklist of the core enterprise, and target users having no amount paid to picked up goods in six months are defined as inferior users or non-performing users, while target users clear of the above circumstances are defined as quality users.

114. The storage medium of any one of claims 92 to 113, wherein the classified type to which each target user corresponds is judged as to whether the target user is a quality user or an inferior user, those target users that pertain to non-performing users are screened out, and the first user data to which these target users correspond is analyzed to obtain corresponding target variables.
Date recue/ date received 2022-02-18

115. The storage medium of any one of claims 92 to 114, wherein the formula for calculating the credit line of the target user is:
credit line = max (basic line * coefficient ¨ Central Bank credit investigation used line, 0)

116. The storage medium of claim 115, wherein the Central Bank credit investigation used line indicates credit extension obtained in a name of the core enterprise, is the used credit line extracted from a credit investigation report of the Central Bank, wherein the basic line is the first line, and the Central Bank credit investigation used line is the second line, wherein the principle of setting credit lines of the target user is that the profitability of the user is capable of covering the existent credit line and the newly added credit line.

117. The storage medium of any one of claims 115 to 116, wherein he basic line takes into consideration the safety of the line as provided, after the user has passed the target admission condition, the basic line is related to the repayment capability of the user, while the repayment capability is related to the income of the user.

118. The storage medium of any one of claims 115 to 117, wherein the coefficient is determined after the basic line has been well determined, wherein the coefficient is a parameter with which to adequately readjust the basic line based on other information or newly appearing information after the basic line has been given, wherein higher the coefficient is, the better the credit of the core enterprise, wherein a lower coefficient indicates that the core enterprise falls short of perfect credit or leaves something to be desired of the credit.

119. The storage medium of any one of claims 115 to 118, wherein after the coefficient has been determined, it is then considered to readjust the coefficient, wherein the readjustment of the coefficient mainly reflects information of the user relevant to recent change in credit risk, and information includes one or more of possible negative early warning signals, abrupt decrease in the sales, amount of money, and overdue circumstances of currently existing credits, wherein the readjustment of the coefficient aims to adequately lower the credit level of the client, control the credit line of the client, and even reject loan to the client when potential risk is encountered.

Date recue/ date received 2022-02-18

120. The storage medium of any one of claims 115 to 119, wherein the main conditions relevant to readjustment of the coefficient comprises:
lowering the coefficient for one grade if there is an early warning signal against the target user;
lowering the coefficient for one grade if the goods picking-up amount of money of the target user decreases by 20% in the recent three months on a year-on-year basis;
lowering the coefficient for one grade if the goods picking-up amount of money of the target user decreases by 15% in last year on a year-on-year basis;
lowering the coefficient for one grade if abnormality occurs to the target user; and lowering the coefficient for one grade if the target user had overdue circumstances in the last two years, with the maximum overdue period being less than 15 days, and all overdue loans have been made good.

121. The storage medium of any one of claims 92 to 120, wherein the first user data of the target user is requested from the first target platform, information of the downstream target users contents as the format of the required data and time duration of the required data can be ascertained.

Date recue/ date received 2022-02-18