A credit quota sharing method, device, terminal and storage medium based on big data

Abstract

The present invention relates to the technical field of big data analysis, and provides a credit line sharing method, device, terminal and storage medium based on big data, including: in response to an applicant's credit line application instruction, obtaining the applicant's user data and application credit line; calculating a first comprehensive credit score based on the user data; using a long short-term memory network model to calculate a risk score based on the user data and the first comprehensive credit score; when it is determined that the risk score is greater than a preset risk score threshold, constructing a social network graph of the applicant; determining multiple target credit line sharing users based on the social network graph; and sending credit line sharing applications to the multiple target credit line sharing users. The present invention can improve the review efficiency of credit line applications, and the security of credit is effectively guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of big data analysis, and in particular to a credit line sharing method, device, terminal and storage medium based on big data. Background Art

[0002] With the rapid development of China's credit market, the credit service market has also been continuously benefited. The credit service market refers to various economic organizations and intermediary institutions that specialize in credit products and consumption to ensure the smooth conduct of market transactions. The technological innovation of Internet credit services has extended the boundaries of credit services, assisted basic credit services, and jointly improved the social credit system. According to the credit score, various service quotas are provided to users, such as deposit mortgage quotas, book rental quotas, credit card quotas, etc., and the application of quotas is becoming more and more extensive. At present, the credit service quotas of users are regularly evaluated, and they can apply for an increase in quotas after using them for a period of time, and the application time is relatively long. For users who use frequently, there is often a phenomenon that the quota is not enough, and it takes a long time to review to increase the quota, but for users who use less, there will be a phenomenon of redundant quotas. In order to quickly increase the quota for users and solve short-term emergency needs, you can use the idle quotas of other users to share.

[0003] Invention patent CN201711006632.9 discloses a credit card credit limit transfer method and system, which determines whether the credit status of the lender's credit card account and the borrower's credit card account meets the requirements; determines whether the transfer period is within the specified period; and determines whether the transfer limit is less than or equal to the remaining limit of the lender's credit card account to review whether it is passed, giving full play to the role of bank credit limit assets and improving bank asset utilization. However, the invention is limited to credit card limits, and too few aspects are considered. It cannot guarantee that users will abide by the contract, resulting in a lack of security and an increase in the risk of bad debts of credit limit assets. Summary of the invention

[0004] In view of the above, it is necessary to propose a credit limit sharing method, device, terminal and storage medium based on big data, which can improve the review efficiency of credit limit applications, effectively guarantee the security of credit, and reduce the risk of bad debts.

[0005] A first aspect of the present invention provides a credit line sharing method based on big data, the method comprising:

[0006] In response to an applicant's quota application instruction, obtain the applicant's user data and application quota;

[0007] Calculating a first comprehensive credit score based on the user data;

[0008] Calculating a risk score based on the user data and the first comprehensive credit score using a long short-term memory network model;

[0009] When it is determined that the risk score is greater than a preset risk score threshold, constructing a social network graph of the applicant;

[0010] Determine a plurality of target quota sharing users based on the social network graph;

[0011] Send a quota sharing application to the multiple target quota sharing users.

[0012] In an optional implementation, the user data includes basic data, credit limit usage data, historical credit data, third-party credit score data, and user social data, and calculating the first comprehensive credit score based on the user data includes:

[0013] Calculate user portrait based on the basic data;

[0014] Predicting the user's spending capacity based on the credit usage data;

[0015] Predicting the user's ability to fulfill a contract based on the historical credit data;

[0016] The first comprehensive credit score is calculated based on the user portrait, the user's consumption ability, the user's contract performance ability, the third-party credit score data and the user's social data.

[0017] In an optional implementation, the calculating the user portrait based on the basic data includes:

[0018] Acquire first data corresponding to the first data label and second data corresponding to the second data label in the basic data;

[0019] Extracting a first feature vector of the first data and a second feature vector of the second data using a BERT model;

[0020] A user portrait is calculated based on the first feature vector and the corresponding first weight, the second feature vector and the corresponding second weight.

[0021] In an optional implementation, the predicting the user's spending capacity based on the credit usage data includes:

[0022] Calculating usage habits, usage frequency, credit limit utilization rate and credit limit overdraft rate based on the credit limit usage data;

[0023] Constructing a user behavior preference vector according to the usage habits, the usage frequency, the credit limit usage rate and the credit limit overdraft rate;

[0024] The trained consumption capacity prediction model is used to predict the user's consumption capacity based on the user behavior preference vector.

[0025] In an optional implementation, the predicting the user's ability to fulfill a contract based on the historical credit data includes:

[0026] Obtaining the fulfillment date and credit card debt in the historical credit data;

[0027] Performing a first binning process on the performance date to obtain a binning date;

[0028] Performing a second binning process on the credit card liabilities to obtain binned liabilities;

[0029] Concatenating the binning date and the binning liability to obtain a user performance vector;

[0030] The trained fulfillment capability prediction model is used to predict the user's fulfillment capability based on the user fulfillment vector.

[0031] In an optional implementation, the calculation process of the third-party credit score data includes:

[0032] receiving credit score data of the applicant sent by a plurality of preset data service structures;

[0033] Allocating scoring weights according to the sending time of the preset data service agency;

[0034] The third-party credit score data is calculated based on the credit score data and the corresponding score weight.

[0035] In an optional implementation, the step of constructing the social network graph of the applicant includes:

[0036] Obtaining multiple contacts of the applicant;

[0037] Calculating the social connection between the applicant and each contact;

[0038] A social network graph is constructed by taking the applicant as the central node in the social network graph, taking the contacts as network nodes, and taking the social association degree as the weight of the edge between the central node and the corresponding network node.

[0039] In an optional implementation, the calculating the social association between the applicant and each contact includes:

[0040] Calculate the social frequency and social time of the applicant and each contact;

[0041] Extracting the social frequency and the social time to construct a social feature vector;

[0042] The social relevance calculation model is used to calculate the social relevance between the applicant and each contact based on the social feature vector.

[0043] In an optional implementation, determining a plurality of target quota sharing users based on the social network graph includes:

[0044] Obtaining the remaining credit limit and second comprehensive credit score of the contact;

[0045] constructing a contact feature vector based on the social association degree, the remaining credit limit and the second comprehensive credit score;

[0046] Calculate the matching degree between the contact and the applicant based on the contact feature vector using a matching degree calculation model;

[0047] According to the matching degree, the social network graph is updated to obtain a target social network graph;

[0048] Search the target social network graph according to preset search rules to obtain multiple target network nodes;

[0049] The multiple target network nodes are determined to be multiple target quota sharing users.

[0050] In an optional implementation, the preset search rule is to start from the central node and search for contacts corresponding to network nodes whose paths are no more than a preset number as the target quota sharing users of the applicant.

[0051] In an optional implementation, after sending the quota sharing application to the multiple target quota sharing users, the method further includes:

[0052] Receiving the amount shared by the target amount sharing user;

[0053] Determine whether the amount shared by any target amount sharing user is equal to the applied amount;

[0054] When the amount shared by any target amount sharing user is equal to the applied amount, receiving the amount shared by any target amount sharing user;

[0055] When the amount shared by any target amount sharing user is less than the applied amount, multiple target amounts are selected from the amounts shared by the multiple target amount sharing users so that the sum of the multiple target amounts can be equal to the applied amount.

[0056] In an optional embodiment, the method further comprises:

[0057] Obtaining a first target matching degree greater than a preset first matching degree threshold in the target social network graph, and providing a credit sharing reservation function to a contact corresponding to the first target matching degree; or

[0058] Obtain a second target matching degree in the target social network graph that is less than a preset second matching degree threshold, display a credit sharing authority option on a network node corresponding to the second target matching degree, receive a confirmed selection from the applicant in the credit sharing authority option, and refuse to receive a credit sharing application from the contact corresponding to the confirmed selection.

[0059] In an optional embodiment, the method further comprises:

[0060] Calculating additional interest and distributing benefits according to the risk score includes: using an additional interest calculation model to calculate the additional interest based on the risk score, the applicant's original credit limit and the applied credit limit; calculating the distributed benefits of the target credit limit sharing users corresponding to the multiple target credit limits according to the additional interest and the second comprehensive credit score.

[0061] In an optional embodiment, the additional interest calculation model is the following formula: Y=(X1+X2) / X3, wherein Y represents the additional interest, X1 represents the risk score, X2 represents the application amount, and X3 represents the original amount.

[0062] In an optional embodiment, the method further comprises:

[0063] Acquiring the credit behavior of the applicant with respect to the credit limit shared by the target credit limit sharing user;

[0064] The first comprehensive credit score of the applicant and the second comprehensive credit score of the target credit limit sharing user are updated according to the credit behavior.

[0065] In an optional embodiment, the method further comprises:

[0066] Obtaining account information of the applicant;

[0067] Determining whether the remaining credit limit in the account information is a preset credit limit threshold;

[0068] When it is determined that the remaining credit limit is the preset credit limit threshold, executing a response to the credit limit application instruction;

[0069] When it is determined that the remaining credit limit is not the preset credit limit threshold, the response to the credit limit application instruction is rejected.

[0070] In an optional embodiment, the method further comprises:

[0071] The target quota sharing user is identified as the applicant's guarantor.

[0072] A second aspect of the present invention provides a credit line sharing device based on big data, the device comprising:

[0073] An application response module, used to respond to an applicant's quota application instruction and obtain the applicant's user data and application quota;

[0074] A first calculation module, configured to calculate a first comprehensive credit score based on the user data;

[0075] A second calculation module, configured to calculate a risk score based on the user data and the first comprehensive credit score using a long short-term memory network model;

[0076] A graph construction module, configured to construct a social network graph of the applicant when it is determined that the risk score is greater than a preset risk score threshold;

[0077] A target determination module, used to determine a plurality of target quota sharing users based on the social network graph;

[0078] The sharing sending module is used to send a quota sharing application to the multiple target quota sharing users.

[0079] A third aspect of the present invention provides a terminal, the terminal comprising a processor, the processor being configured to implement the credit line sharing method based on big data when executing a computer program stored in a memory.

[0080] A fourth aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, and the computer program, when executed by a processor, implements the credit line sharing method based on big data.

[0081] In summary, the credit line sharing method, device, terminal and storage medium based on big data described in the present invention, upon receiving the credit line application instruction of the applicant, obtain the user data and application credit line of the applicant; calculate the first comprehensive credit score based on the user data, and then use the long short-term memory network model to calculate the risk score based on the user data and the first comprehensive credit score; and when it is determined that the risk score is greater than the preset risk score threshold, construct the social network map of the applicant; determine multiple target credit line sharing users based on the social network map; send credit line sharing applications to the multiple target credit line sharing users. It can solve the technical problems of credit line redundancy and long review time for credit line increase in existing credit services, and proposes that credit line can be shared between different users in the same platform, which reduces the review time for credit line increase applications, improves the review efficiency of credit line increase applications, and through risk assessment and score update, the security of credit is effectively guaranteed and the risk of bad debts is reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0082] Figure 1 It is a flow chart of a credit quota sharing method based on big data provided in Example 1 of the present invention.

[0083] Figure 2 It is a structural diagram of a credit limit sharing device based on big data provided in the second embodiment of the present invention.

[0084] Figure 3 It is a schematic diagram of the structure of a social network graph provided by an embodiment of the present invention.

[0085] Figure 4 It is a schematic diagram of the structure of a terminal provided in Embodiment 3 of the present invention. DETAILED DESCRIPTION

[0086] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present invention and the features in the embodiments can be combined with each other without conflict.

[0087] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art of the present invention. The terms used in the specification of the present invention herein are only for the purpose of describing specific embodiments and are not intended to limit the present invention.

[0088] The credit line sharing method based on big data provided in the embodiment of the present invention is executed by a terminal, and accordingly, the credit line sharing device based on big data runs in the terminal.

[0089] Figure 1This is a flow chart of a credit line sharing method based on big data provided by Embodiment 1 of the present invention. The credit line sharing method based on big data specifically includes the following steps. According to different requirements, the order of the steps in the flow chart can be changed, and some steps can be omitted.

[0090] S11, in response to the quota application instruction of the applicant, obtaining the user data and application quota of the applicant.

[0091] The terminal may provide a credit application user interface, and the applicant may log in to the credit application user interface and touch the credit application virtual icon on the credit application user interface to trigger the credit application instruction. The terminal responds to the applicant's credit application instruction and obtains the applicant's user data and application credit.

[0092] Among them, the applicant can be the user who needs to increase the credit limit, or can share the credit limit with other users who need to increase the credit limit.

[0093] The user data may include, but is not limited to: basic data. Basic data may include: gender, age, marital status, education, occupation, income flow, bank card status, social security and other objective data that can reflect the applicant's personal credit information.

[0094] In an optional embodiment, the method further comprises:

[0095] Obtaining account information of the applicant;

[0096] Determining whether the remaining credit limit in the account information is a preset credit limit threshold;

[0097] When it is determined that the remaining credit limit is the preset credit limit threshold, executing a response to the credit limit application instruction;

[0098] When it is determined that the remaining credit limit is not the preset credit limit threshold, the response to the credit limit application instruction is rejected.

[0099] When receiving the credit application instruction from the applicant, the terminal obtains the account information of the applicant, wherein the account information may include: account name, credit limit per cycle, and remaining credit limit.

[0100] The terminal determines whether the credit limit of the current cycle in the applicant's account information has been exhausted, that is, whether the remaining credit limit is a preset limit threshold (for example, 0). If the credit limit of the applicant's current cycle has been exhausted, that is, the remaining credit limit is the preset limit threshold, then respond to the applicant's credit limit application instruction to determine the credit limit applied by the applicant. If the credit limit of the applicant's current cycle has not been exhausted, that is, the remaining credit limit is not the preset limit threshold, then the response to the credit limit application instruction is rejected. Exemplarily, taking the credit limit of a credit card as an example, the credit limit of a credit card is generally calculated on a monthly basis. Assuming that the applicant's monthly credit card limit is 3,000, when the remaining credit limit of the credit card in that month is 0, it indicates that the applicant's credit limit of the current cycle has been exhausted, and the terminal responds to the credit limit application instruction. In this way, it is possible to prevent the flooding of credit limit applications and avoid users from applying for credit limits at will, resulting in a waste of resources.

[0101] S12: Calculate a first comprehensive credit score based on the user data.

[0102] In an optional implementation, in addition to the above basic data, user data also includes credit limit usage data, historical credit data, third-party credit score data, and user social data. Basic data, credit limit usage data, historical credit data, third-party credit score data, and user social data provide a data basis for calculating the user's credit score.

[0103] In an optional implementation, the calculating a first comprehensive credit score based on the user data includes:

[0104] Calculate user portrait based on the basic data;

[0105] Predicting the user's spending capacity based on the credit usage data;

[0106] Predicting the user's ability to fulfill a contract based on the historical credit data;

[0107] The first comprehensive credit score is calculated based on the user portrait, the user's consumption ability, the user's contract performance ability, the third-party credit score data and the user's social data.

[0108] After calculating the user portrait and predicting the user's consumption capacity and the user's ability to fulfill contracts, the terminal adds and averages the user portrait, the user's consumption capacity, the user's ability to fulfill contracts, the third-party credit score data and the user's social data to obtain the applicant's first comprehensive credit score.

[0109] In an optional implementation, the calculating the user portrait based on the basic data includes:

[0110] Acquire first data corresponding to the first data label and second data corresponding to the second data label in the basic data;

[0111] Extracting a first feature vector of the first data and a second feature vector of the second data using a BERT model;

[0112] A user portrait is calculated based on the first feature vector and the corresponding first weight, the second feature vector and the corresponding second weight.

[0113] Each data in the basic data is marked with a data label, some data are marked with a first data label, and some data are marked with a second data label. For example, the education level, occupation, income flow, bank card status, social security, etc. in the basic data are marked with a first data label, and the gender, age, marital status, etc. in the basic data are marked with a second data label.

[0114] The BRET (Bidirectional Encoder Representations from Transformer) model is a pre-trained language model. The BRET model can process each character in the data and obtain the feature vector of each character.

[0115] Since the data marked with the first data label, such as education, occupation, income flow, bank card status, social security, etc., can strongly reflect the personal credit information of the applicant, the data marked with the first data label is determined as strongly related data, while the data marked with the second data label, such as gender, age, marital status, etc., cannot strongly reflect the personal credit information of the applicant, so the data marked with the second data label is determined as weakly related data. A first weight is configured for the first feature vector, and a second weight is configured for the second feature vector, wherein the first weight is greater than the second weight. A first product between the first feature vector and the corresponding first weight is calculated, a second product between the second feature vector and the corresponding second weight is calculated, and the sum of the products between the first product and the second product is calculated to obtain a user portrait.

[0116] This optional implementation, by identifying different data tags for the data, configuring different weights based on different data tags, and calculating user portraits based on the data identified with different data tags and the corresponding weights, can focus on people who have the ability to repay but do not have much credit history, such as recent graduates. Therefore, it is highly practical and has broad application prospects.

[0117] In an optional implementation, the predicting the user's spending capacity based on the credit usage data includes:

[0118] Calculating usage habits, usage frequency, credit limit utilization rate and credit limit overdraft rate based on the credit limit usage data;

[0119] Constructing a user behavior preference vector according to the usage habits, the usage frequency, the credit limit usage rate and the credit limit overdraft rate;

[0120] The trained consumption capacity prediction model is used to predict the user's consumption capacity based on the user behavior preference vector.

[0121] The terminal obtains the date of use, amount of use, and category of use in the credit usage data, and can calculate the frequency of use based on the date of use, determine the usage habits based on the category of use, and calculate the credit usage rate and credit overdraft rate based on the amount of use, and then splice the usage habits, the frequency of use, the credit usage rate, and the credit overdraft rate to obtain a user behavior preference vector. The user behavior preference vector is input into the pre-trained consumption capacity prediction model, and the consumption capacity of the user is predicted by the consumption capacity prediction model, thereby evaluating the user's future usage potential.

[0122] In an optional implementation, the predicting the user's ability to fulfill a contract based on the historical credit data includes:

[0123] Obtaining the fulfillment date and credit card debt in the historical credit data;

[0124] Performing a first binning process on the performance date to obtain a binning date;

[0125] Performing a second binning process on the credit card liabilities to obtain binned liabilities;

[0126] Concatenating the binning date and the binning liability to obtain a user performance vector;

[0127] The trained fulfillment capability prediction model is used to predict the user's fulfillment capability based on the user fulfillment vector.

[0128] The terminal obtains the fulfillment date and credit card debt in the historical credit data, and can perform equal frequency binning on the fulfillment date to obtain binning date, and can perform card square binning on the credit card debt to obtain binning debt, and then splice the binning date and the binning debt together to obtain the user fulfillment vector. The user fulfillment vector is input into the pre-trained fulfillment ability prediction, and the user's fulfillment ability is predicted through the fulfillment ability prediction, so as to evaluate the user's future repayment potential.

[0129] In an optional implementation, the calculation process of the third-party credit score data includes:

[0130] receiving credit score data of the applicant sent by a plurality of preset data service structures;

[0131] Allocating scoring weights according to the sending time of the preset data service agency;

[0132] The third-party credit score data is calculated based on the credit score data and the corresponding score weight.

[0133] The preset data service structure may be, for example, Sesame Credit, Tencent Credit, Orange Credit, etc. The terminal may pre-set multiple data service agencies, and in response to the applicant's credit application instruction, send the applicant's credit score data request to the preset data service structure, so that the preset data service structure sends the applicant's credit score data to the terminal. Since different data service agencies have different response speeds when sending the applicant's credit score data, it is generally believed that the faster the response speed of the data service agency, the more authoritative it is, and the slower the response speed of the data service agency, the less authoritative it is. Different scoring weights are assigned according to the sending time of the preset data service agency, and the third-party credit score data is obtained by weighted calculation based on the credit score data and the corresponding scoring weights.

[0134] Among them, user social data may include: social activity, social density and other information. User social data can evaluate the user's social circle and interpersonal network.

[0135] S13, using a long short-term memory network model to calculate a risk score based on the user data and the first comprehensive credit score.

[0136] For credit service products, credit risks must be strictly controlled to avoid arrears that would cause losses to the interests of credit service institutions.

[0137] The terminal trains the long short-term memory network model in advance based on other applicants’ basic data, credit limit usage data, historical credit data, third-party credit score data, user social data and credit score.

[0138] The user data and the credit score calculation are input into the trained long short-term memory network model, and the risk score is output through the long short-term memory network model.

[0139] In an optional embodiment, the method further comprises:

[0140] comparing the risk score to a risk score threshold;

[0141] When it is determined that the risk score is less than or equal to the preset risk score threshold, a credit limit sharing application instruction is triggered;

[0142] When it is determined that the risk score is greater than the preset risk score threshold, triggering the quota sharing application instruction is prohibited.

[0143] The terminal determines the risk of the applicant's credit application by comparing the risk score with the preset risk score threshold, thereby determining whether the applicant can repay the credit on time. When it is determined that the risk score is less than or equal to the preset risk score threshold, it indicates that the applicant's credit application risk is low, and the credit sharing application instruction is triggered to allow the applicant to seek credit sharing from others, thereby increasing the applicant's credit. When it is determined that the risk score is greater than the preset risk score threshold, it indicates that the applicant's credit application risk is high, and the credit sharing application instruction is prohibited from being triggered to prohibit the applicant from seeking credit sharing from others, thereby protecting the credit of others.

[0144] For example, assuming that the applicant's monthly credit card limit is 3,000, the remaining credit card limit for this cycle is 0, and the applicant applies for a limit of 1,000. Since the risk score obtained through the long short-term memory network model is greater than the preset risk score threshold, the credit sharing application instruction is prohibited from being triggered to protect the interests of other people and credit service agencies, such as shared bicycles, libraries, banks and other institutions. If the applicant has good credit behavior, low overdraft rate, high income, and suddenly needs credit turnover, the risk score obtained through the long short-term memory network model is less than or equal to the preset risk score threshold, then the credit sharing application instruction is triggered, so that the user can apply for credit sharing to solve the applicant's credit turnover problem. In this way, there is no need to wait for the long review time of the credit increase application, and the credit service agency also obtains the benefits of the corresponding credit being used, reducing the phenomenon of market credit redundancy and improving resource utilization.

[0145] S14: When it is determined that the risk score is greater than a preset risk score threshold, construct a social network graph of the applicant.

[0146] The terminal constructs a social network graph centered on the applicant for the applicant, and determines multiple target credit line sharing users based on the social network graph, thereby sending credit line sharing invitations to the multiple target credit line sharing users.

[0147] In an optional implementation, the step of constructing the social network graph of the applicant includes:

[0148] Obtaining multiple contacts of the applicant;

[0149] Calculating the social connection between the applicant and each contact;

[0150] A social network graph is constructed by taking the applicant as the central node in the social network graph, taking the contacts as network nodes, and taking the social association degree as the weight of the edge between the central node and the corresponding network node.

[0151] like Figure 3 As shown, the applicant is regarded as the central node in the social network graph, and each contact is regarded as a network node in the social network graph. An undirected edge is established between the central node and each network node, and the weight of the edge is the social association between the central node and the network node.

[0152] In an optional implementation, the calculating the social association between the applicant and each contact includes:

[0153] Calculate the social frequency and social time of the applicant and each contact;

[0154] Extracting the social frequency and the social time to construct a social feature vector;

[0155] The social relevance calculation model is used to calculate the social relevance between the applicant and each contact based on the social feature vector.

[0156] The terminal can obtain the social frequency and social time of the applicant and each contact through the operator's call records, communication records of communication software, etc., thereby constructing a social feature vector based on the social frequency and the social time, and based on the social feature vector, using a pre-trained social association calculation model to calculate the social association between the applicant and each contact.

[0157] S15, determining a plurality of target quota sharing users based on the social network graph.

[0158] In an optional implementation, determining a plurality of target quota sharing users based on the social network graph includes:

[0159] Obtaining the remaining credit limit and second comprehensive credit score of the contact;

[0160] constructing a contact feature vector based on the social association degree, the remaining credit limit and the second comprehensive credit score;

[0161] Calculate the matching degree between the contact and the applicant based on the contact feature vector using a matching degree calculation model;

[0162] According to the matching degree, the social network graph is updated to obtain a target social network graph;

[0163] Search the target social network graph according to preset search rules to obtain multiple target network nodes;

[0164] The multiple target network nodes are determined to be multiple target quota sharing users.

[0165] The terminal may obtain the user data of the contact and calculate a second comprehensive credit score based on the user data of the contact. The process of calculating the second comprehensive credit score is the same as the process of calculating the first comprehensive credit score. For details, please refer to the above step S12 and its related description.

[0166] The terminal may pre-train a matching degree calculation model, input a contact feature vector into the trained matching degree calculation model, and output a matching degree through the trained matching degree calculation model as the matching degree between the corresponding contact and the applicant.

[0167] By replacing the social association degree in the social network graph with the matching degree, the social network graph is updated to obtain the target social network graph, in which the weight of the edge between the central node and the network node in the target social network graph is the matching degree.

[0168] In an optional implementation, the preset search rule may be to start from the central node and search for contacts corresponding to network nodes whose paths are no more than a preset number (eg, 5) as the applicant's target quota sharing users.

[0169] S16: Send a quota sharing application to the multiple target quota sharing users.

[0170] The target credit limit sharing users searched out are contacts with large remaining credit limits, high social connections, and high second comprehensive credit scores. These contacts are more willing to accept credit limit sharing applications.

[0171] The terminal may select the target quota sharing user who accepts the quota sharing application the fastest as the sharer.

[0172] In an optional implementation, the terminal may also determine the target credit limit sharing user as the applicant's guarantor, which can effectively solve the technical problem of no one to provide a guarantor in the event of an applicant's default.

[0173] In an optional implementation, after sending the quota sharing application to the multiple target quota sharing users, the method further includes:

[0174] Receiving the amount shared by the target amount sharing user;

[0175] Determine whether the amount shared by any target amount sharing user is equal to the applied amount;

[0176] When the amount shared by any target amount sharing user is equal to the applied amount, receiving the amount shared by any target amount sharing user;

[0177] When the amount shared by any target amount sharing user is less than the applied amount, multiple target amounts are selected from the amounts shared by the multiple target amount sharing users so that the sum of the multiple target amounts can be equal to the applied amount.

[0178] It should be understood that the amount shared by the target amount sharing user will not be greater than the applied amount.

[0179] When the amount shared by a target quota sharing user is equal to the application amount, the amount shared by the target quota sharing user equal to the application amount can be directly received. When the amount shared by any target quota sharing user is less than the application amount, indicating that the remaining amount of all target quota sharing users is insufficient to share with the applicant, the terminal allocates the amount of multiple target quota sharing users, so that the target quota sharing users corresponding to the selected multiple target quotas can jointly complete the applicant's quota sharing application.

[0180] In an optional embodiment, the method further includes: obtaining a first target matching degree in the target social network graph that is greater than a preset first matching degree threshold, and providing a credit sharing reservation function to contacts corresponding to the first target matching degree.

[0181] In actual applications, when the applicant's credit limit is high and often unused, one or more contacts will invite him to share credit limit. The terminal can allow the applicant's contacts to make credit limit reservations by providing a credit limit sharing reservation function, so that the applicant can plan his credit limit in advance. Since the first contact corresponding to the first target matching degree in the target social network graph that is greater than the preset first matching degree threshold may have a very close relationship with the applicant, in order to be able to send credit limit sharing to the first contact, the terminal can provide a credit limit sharing reservation function to the contact corresponding to the first target matching degree, so that the first contact corresponding to the first target matching degree can send a credit limit sharing application to the applicant as soon as possible, so that the applicant can reserve credit limit for the first contact.

[0182] In an optional embodiment, the method also includes: obtaining a second target matching degree in the target social network graph that is less than a preset second matching degree threshold, displaying a credit sharing authority option on a network node corresponding to the second target matching degree, receiving a confirmed selection from the applicant in the credit sharing authority option, and refusing to receive a credit sharing application from the contact corresponding to the confirmed selection.

[0183] Wherein, the preset first matching degree threshold is greater than the preset second matching degree threshold.

[0184] The second contact corresponding to the second target matching degree in the target social network graph that is less than the preset second matching degree threshold is not very close to the applicant. In order to avoid receiving the credit sharing application from the second contact and disturbing the applicant, the credit sharing permission option is displayed on the network node corresponding to the second target matching degree in the target social network graph, so that the applicant can choose whether to reject the credit sharing application sent by the second contact. When the applicant clicks on the credit sharing permission option, the terminal receives the applicant's confirmation selection instruction in the credit sharing permission option, so that when receiving the credit sharing application of the contact corresponding to the confirmation selection, the terminal directly intercepts the credit sharing application of the contact corresponding to the confirmation selection to refuse to receive the credit sharing application of the contact corresponding to the confirmation selection.

[0185] In an optional embodiment, the method further comprises:

[0186] Additional interest calculation and profit distribution are performed based on the risk score.

[0187] In order to solve the problem of overdraft and urgent need for credit turnover, the applicant should pay additional interest as the handling fee of the information service agency and the profit of the beneficiary, so that all three parties can benefit and achieve three goals at one stroke.

[0188] In an optional implementation, the additional interest calculation and income distribution according to the risk score includes:

[0189] Calculating the additional interest using an additional interest calculation model based on the risk score, the original credit limit of the applicant, and the application limit;

[0190] The distributed income of the target credit limit sharing users corresponding to the multiple target credit limits is calculated based on the additional interest and the second comprehensive credit score.

[0191] The additional interest calculation model can be the following formula: Y = (X1 + X2) / X3, where Y represents the additional interest, X1 represents the risk score, X2 represents the application amount, and X3 represents the original amount. It can be seen that the higher the risk score, the larger the application amount, the smaller the original amount, and the more additional interest; the lower the risk score, the smaller the application amount, the larger the original amount, and the less additional interest.

[0192] The terminal calculates the credit score sum of the second comprehensive credit scores of the target credit sharing users corresponding to the multiple target credits, and then calculates the ratio of the second comprehensive credit score of the target credit sharing user corresponding to each target credit to the credit score sum, and obtains the income of the corresponding target credit sharing user according to the product of the ratio and the additional interest. The higher the second comprehensive credit score, the larger the corresponding ratio, and the higher the income of the corresponding target credit sharing user; the lower the second comprehensive credit score, the smaller the corresponding ratio, and the lower the income of the corresponding target credit sharing user.

[0193] In this optional implementation, additional interest calculation and profit distribution are performed according to the risk score, which can increase the user's willingness to share the credit limit, motivate the user's trustworthy behavior, improve the user's credit score, and improve the credit service market.

[0194] The terminal may also share the user's earnings according to the product of the ratio and the preset ratio of the additional interest as the corresponding target amount.

[0195] In an optional embodiment, the method further comprises:

[0196] Acquiring the credit behavior of the applicant with respect to the credit limit shared by the target credit limit sharing user;

[0197] The first comprehensive credit score of the applicant and the second comprehensive credit score of the target credit limit sharing user are updated according to the credit behavior.

[0198] After the target credit sharing user shares the credit with the applicant, it indicates that the applicant has successfully applied for the credit, and the terminal obtains the credit behavior of the applicant thereafter, which may include whether the books are returned on time or whether the loan is repaid on time. If the applicant returns the books or repays the loan on time, the credit behavior is considered to be trustworthy, and if the applicant fails to return the books or repay the loan on time, the credit behavior is considered to be untrustworthy.

[0199] The trustworthy behavior of the applicant also serves as the trustworthy behavior of the target credit sharing user, and the first comprehensive credit score of the applicant and the second comprehensive credit score of the target credit sharing user are updated respectively, and the social association between the applicant and the target credit sharing user is updated, thereby updating the matching degree between the applicant and the target credit sharing user, and then updating the social network map of the applicant. In this way, the target credit sharing user can guarantee the credit of the applicant, promote the trustworthy behavior of the applicant, and protect the interests of the credit service agency.

[0200] In an optional embodiment, the terminal can also store the credit sharing. When the applicant sends a second credit sharing application to the target credit sharing user again, the weight of the edge in the social network graph of the applicant is increased, thereby making the applicant's second credit sharing application more likely to succeed, reducing the computational complexity, and at the same time reducing the risk score.

[0201] Figure 2 It is a structural diagram of a credit limit sharing device based on big data provided in the second embodiment of the present invention.

[0202] In some embodiments, the credit line sharing device 20 based on big data may include a plurality of functional modules composed of computer program segments. The computer programs of the various program segments in the credit line sharing device 20 based on big data may be stored in the memory of the terminal and executed by at least one processor to execute (see Figure 1 Description) Credit quota sharing function based on big data.

[0203] In this embodiment, the credit line sharing device 20 based on big data can be divided into multiple functional modules according to the functions it performs. The functional modules may include: an application response module 201, a first calculation module 202, a second calculation module 203, a graph construction module 204, a target determination module 205, a sharing and sending module 206, a credit line receiving module 207, a function providing module 208, an interest distribution module 209 and a scoring update module 210. The module referred to in the present invention refers to a series of computer program segments that can be executed by at least one processor and can complete fixed functions, which are stored in a memory. In this embodiment, the functions of each module will be described in detail in subsequent embodiments.

[0204] The application response module 201 is used to respond to the quota application instruction of the applicant and obtain the user data and application quota of the applicant.

[0205] The terminal may provide a credit application user interface, and the applicant may log in to the credit application user interface and touch the credit application virtual icon on the credit application user interface to trigger the credit application instruction. The terminal responds to the applicant's credit application instruction and obtains the applicant's user data and application credit.

[0206] Among them, the applicant can be the user who needs to increase the credit limit, or can share the credit limit with other users who need to increase the credit limit.

[0207] The user data may include, but is not limited to: basic data. Basic data may include: gender, age, marital status, education, occupation, income flow, bank card status, social security and other objective data that can reflect the applicant's personal credit information.

[0208] In an optional implementation, the application response module 201 is further configured to:

[0209] Obtaining account information of the applicant;

[0210] Determining whether the remaining credit limit in the account information is a preset credit limit threshold;

[0211] When it is determined that the remaining credit limit is the preset credit limit threshold, executing a response to the credit limit application instruction;

[0212] When it is determined that the remaining credit limit is not the preset credit limit threshold, the response to the credit limit application instruction is rejected.

[0213] When receiving the credit application instruction from the applicant, the terminal obtains the account information of the applicant, wherein the account information may include: account name, credit limit per cycle, and remaining credit limit.

[0214] The terminal determines whether the credit limit of the current cycle in the applicant's account information has been exhausted, that is, whether the remaining credit limit is a preset limit threshold (for example, 0). If the credit limit of the applicant's current cycle has been exhausted, that is, the remaining credit limit is the preset limit threshold, then respond to the applicant's credit limit application instruction to determine the credit limit applied by the applicant. If the credit limit of the applicant's current cycle has not been exhausted, that is, the remaining credit limit is not the preset limit threshold, then the response to the credit limit application instruction is rejected. Exemplarily, taking the credit limit of a credit card as an example, the credit limit of a credit card is generally calculated on a monthly basis. Assuming that the applicant's monthly credit card limit is 3,000, when the remaining credit limit of the credit card in that month is 0, it indicates that the applicant's credit limit of the current cycle has been exhausted, and the terminal responds to the credit limit application instruction. In this way, it is possible to prevent the flooding of credit limit applications and avoid users from applying for credit limits at will, resulting in a waste of resources.

[0215] The first calculation module 202 is used to calculate a first comprehensive credit score based on the user data.

[0216] In an optional implementation, in addition to the above basic data, user data also includes credit limit usage data, historical credit data, third-party credit score data, and user social data. Basic data, credit limit usage data, historical credit data, third-party credit score data, and user social data provide a data basis for calculating the user's credit score.

[0217] In an optional implementation, the first calculation module 202 calculates the first comprehensive credit score based on the user data including:

[0218] Calculate user portrait based on the basic data;

[0219] Predicting the user's spending capacity based on the credit usage data;

[0220] Predicting the user's ability to fulfill a contract based on the historical credit data;

[0221] The first comprehensive credit score is calculated based on the user portrait, the user's consumption ability, the user's contract performance ability, the third-party credit score data and the user's social data.

[0222] After calculating the user portrait and predicting the user's consumption capacity and the user's ability to fulfill contracts, the terminal adds and averages the user portrait, the user's consumption capacity, the user's ability to fulfill contracts, the third-party credit score data and the user's social data to obtain the applicant's first comprehensive credit score.

[0223] In an optional implementation, the calculating the user portrait based on the basic data includes:

[0224] Acquire first data corresponding to the first data label and second data corresponding to the second data label in the basic data;

[0225] Extracting a first feature vector of the first data and a second feature vector of the second data using a BERT pre-trained model;

[0226] A user portrait is calculated based on the first feature vector and the corresponding first weight, the second feature vector and the corresponding second weight.

[0227] Each data in the basic data is marked with a data label, some data are marked with a first data label, and some data are marked with a second data label. For example, the education level, occupation, income flow, bank card status, social security, etc. in the basic data are marked with a first data label, and the gender, age, marital status, etc. in the basic data are marked with a second data label.

[0228] The BERT (Bidirectional Encoder Representations from Transformer) model is a pre-trained language model. The BERT model can process each character in the data and obtain the feature vector of each character. Similarly, the BERT pre-trained model can also be replaced by the GPT3 model.

[0229] Since the data marked with the first data label, such as education, occupation, income flow, bank card status, social security, etc., can strongly reflect the personal credit information of the applicant, the data marked with the first data label is determined as strongly related data, while the data marked with the second data label, such as gender, age, marital status, etc., cannot strongly reflect the personal credit information of the applicant, so the data marked with the second data label is determined as weakly related data. A first weight is configured for the first feature vector, and a second weight is configured for the second feature vector, wherein the first weight is greater than the second weight. A first product between the first feature vector and the corresponding first weight is calculated, a second product between the second feature vector and the corresponding second weight is calculated, and the sum of the products between the first product and the second product is calculated to obtain a user portrait.

[0230] This optional implementation, by identifying different data tags for the data, configuring different weights based on different data tags, and calculating user portraits based on the data identified with different data tags and the corresponding weights, can focus on people who have the ability to repay but do not have much credit history, such as recent graduates. Therefore, it is highly practical and has broad application prospects.

[0231] In an optional implementation, the predicting the user's spending capacity based on the credit usage data includes:

[0232] Calculating usage habits, usage frequency, credit limit utilization rate and credit limit overdraft rate based on the credit limit usage data;

[0233] Constructing a user behavior preference vector according to the usage habits, the usage frequency, the credit limit usage rate and the credit limit overdraft rate;

[0234] The trained consumption capacity prediction model is used to predict the user's consumption capacity based on the user behavior preference vector.

[0235] The terminal obtains the date of use, amount of use, and category of use in the credit usage data, and can calculate the frequency of use based on the date of use, determine the usage habits based on the category of use, and calculate the credit usage rate and credit overdraft rate based on the amount of use, and then splice the usage habits, the frequency of use, the credit usage rate, and the credit overdraft rate to obtain a user behavior preference vector. The user behavior preference vector is input into the pre-trained consumption capacity prediction model, and the consumption capacity of the user is predicted by the consumption capacity prediction model, thereby evaluating the user's future usage potential.

[0236] In an optional implementation, the predicting the user's ability to fulfill a contract based on the historical credit data includes:

[0237] Obtaining the fulfillment date and credit card debt in the historical credit data;

[0238] Performing a first binning process on the performance date to obtain a binning date;

[0239] Performing a second binning process on the credit card liabilities to obtain binned liabilities;

[0240] Concatenating the binning date and the binning liability to obtain a user performance vector;

[0241] The trained fulfillment capability prediction model is used to predict the user's fulfillment capability based on the user fulfillment vector.

[0242] The terminal obtains the fulfillment date and credit card debt in the historical credit data, and can perform equal frequency binning on the fulfillment date to obtain binning date, and can perform card square binning on the credit card debt to obtain binning debt, and then splice the binning date and the binning debt together to obtain the user fulfillment vector. The user fulfillment vector is input into the pre-trained fulfillment ability prediction, and the user's fulfillment ability is predicted through the fulfillment ability prediction, so as to evaluate the user's future repayment potential.

[0243] In an optional implementation, the calculation process of the third-party credit score data includes:

[0244] receiving credit score data of the applicant sent by a plurality of preset data service structures;

[0245] Allocating scoring weights according to the sending time of the preset data service agency;

[0246] The third-party credit score data is calculated based on the credit score data and the corresponding score weight.

[0247] The preset data service structure may be, for example, Sesame Credit, Tencent Credit, Orange Credit, etc. The terminal may pre-set multiple data service agencies, and in response to the applicant's credit application instruction, send the applicant's credit score data request to the preset data service structure, so that the preset data service structure sends the applicant's credit score data to the terminal. Since different data service agencies have different response speeds when sending the applicant's credit score data, it is generally believed that the faster the response speed of the data service agency, the more authoritative it is, and the slower the response speed of the data service agency, the less authoritative it is. Different scoring weights are assigned according to the sending time of the preset data service agency, and the third-party credit score data is obtained by weighted calculation based on the credit score data and the corresponding scoring weights.

[0248] Among them, user social data may include: social activity, social density and other information. User social data can evaluate the user's social circle and interpersonal network.

[0249] The second calculation module 203 is used to calculate the risk score based on the user data and the first comprehensive credit score using a long short-term memory network model.

[0250] For credit service products, credit risks must be strictly controlled to avoid arrears that would cause losses to the interests of credit service institutions.

[0251] The terminal trains the long short-term memory network model in advance based on other applicants’ basic data, credit limit usage data, historical credit data, third-party credit score data, user social data and credit score.

[0252] The user data and the credit score calculation are input into the trained long short-term memory network model, and the risk score is output through the long short-term memory network model.

[0253] In an optional implementation, the second calculation module 203 is further configured to:

[0254] comparing the risk score to a risk score threshold;

[0255] When it is determined that the risk score is less than or equal to the preset risk score threshold, a credit limit sharing application instruction is triggered;

[0256] When it is determined that the risk score is greater than the preset risk score threshold, triggering the quota sharing application instruction is prohibited.

[0257] The terminal determines the risk of the applicant's credit application by comparing the risk score with the preset risk score threshold, thereby determining whether the applicant can repay the credit on time. When it is determined that the risk score is less than or equal to the preset risk score threshold, it indicates that the applicant's credit application risk is low, and the credit sharing application instruction is triggered to allow the applicant to seek credit sharing from others, thereby increasing the applicant's credit. When it is determined that the risk score is greater than the preset risk score threshold, it indicates that the applicant's credit application risk is high, and the credit sharing application instruction is prohibited from being triggered to prohibit the applicant from seeking credit sharing from others, thereby protecting the credit of others.

[0258] For example, assuming that the applicant's monthly credit card limit is 3,000, the remaining credit card limit for this cycle is 0, and the applicant applies for a limit of 1,000. Since the risk score obtained through the long short-term memory network model is greater than the preset risk score threshold, the credit sharing application instruction is prohibited from being triggered to protect the interests of other people and credit service agencies, such as shared bicycles, libraries, banks and other institutions. If the applicant has good credit behavior, low overdraft rate, high income, and suddenly needs credit turnover, the risk score obtained through the long short-term memory network model is less than or equal to the preset risk score threshold, then the credit sharing application instruction is triggered, so that the user can apply for credit sharing to solve the applicant's credit turnover problem. In this way, there is no need to wait for the long review time of the credit increase application, and the credit service agency also obtains the benefits of the corresponding credit being used, reducing the phenomenon of market credit redundancy and improving resource utilization.

[0259] The graph construction module 204 is used to construct a social network graph of the applicant when it is determined that the risk score is greater than a preset risk score threshold.

[0260] The terminal constructs a social network graph centered on the applicant for the applicant, and determines multiple target credit line sharing users based on the social network graph, thereby sending credit line sharing invitations to the multiple target credit line sharing users.

[0261] In an optional implementation, the graph construction module 204 constructs the social network graph of the applicant including:

[0262] Obtaining multiple contacts of the applicant;

[0263] Calculating the social connection between the applicant and each contact;

[0264] A social network graph is constructed by taking the applicant as the central node in the social network graph, taking the contacts as network nodes, and taking the social association degree as the weight of the edge between the central node and the corresponding network node.

[0265] like Figure 3 As shown, the applicant is regarded as the central node in the social network graph, and each contact is regarded as a network node in the social network graph. An undirected edge is established between the central node and each network node, and the weight of the edge is the social association between the central node and the network node.

[0266] In an optional implementation, the calculating the social association between the applicant and each contact includes:

[0267] Calculate the social frequency and social time of the applicant and each contact;

[0268] Extracting the social frequency and the social time to construct a social feature vector;

[0269] The social relevance calculation model is used to calculate the social relevance between the applicant and each contact based on the social feature vector.

[0270] The terminal can obtain the social frequency and social time of the applicant and each contact through the operator's call records, communication records of communication software, etc., thereby constructing a social feature vector based on the social frequency and the social time, and based on the social feature vector, using a pre-trained social association calculation model to calculate the social association between the applicant and each contact.

[0271] The target determination module 205 is used to determine a plurality of target quota sharing users based on the social network graph.

[0272] In an optional implementation, determining a plurality of target quota sharing users based on the social network graph includes:

[0273] Obtaining the remaining credit limit and second comprehensive credit score of the contact;

[0274] constructing a contact feature vector based on the social association degree, the remaining credit limit and the second comprehensive credit score;

[0275] Calculate the matching degree between the contact and the applicant based on the contact feature vector using a matching degree calculation model;

[0276] According to the matching degree, the social network graph is updated to obtain a target social network graph;

[0277] Search the target social network graph according to preset search rules to obtain multiple target network nodes;

[0278] The multiple target network nodes are determined to be multiple target quota sharing users.

[0279] The terminal may obtain the user data of the contact and calculate a second comprehensive credit score based on the user data of the contact. The process of calculating the second comprehensive credit score is the same as the process of calculating the first comprehensive credit score. For details, please refer to the above step S12 and its related description.

[0280] The terminal may pre-train a matching degree calculation model, input a contact feature vector into the trained matching degree calculation model, and output a matching degree through the trained matching degree calculation model as the matching degree between the corresponding contact and the applicant.

[0281] By replacing the social association degree in the social network graph with the matching degree, the social network graph is updated to obtain the target social network graph, in which the weight of the edge between the central node and the network node in the target social network graph is the matching degree.

[0282] In an optional implementation, the preset search rule may be to start from the central node and search for contacts corresponding to network nodes whose paths are no more than a preset number (eg, 5) as the applicant's target quota sharing users.

[0283] The sharing sending module 206 is used to send a quota sharing application to the multiple target quota sharing users.

[0284] The target credit limit sharing users searched out are contacts with large remaining credit limits, high social connections, and high second comprehensive credit scores. These contacts are more willing to accept credit limit sharing applications.

[0285] The terminal may select the target quota sharing user who accepts the quota sharing application the fastest as the sharer.

[0286] In an optional implementation, the terminal may also determine the target credit limit sharing user as the applicant's guarantor, which can effectively solve the technical problem of no one to provide a guarantor in the event of an applicant's default.

[0287] In an optional embodiment, after sending a quota sharing application to the multiple target quota sharing users, the quota receiving module 207 is used to: receive the quota shared by the target quota sharing users; determine whether the quota shared by any target quota sharing user is equal to the application quota; when the quota shared by any target quota sharing user is equal to the application quota, receive the quota shared by any target quota sharing user; when the quota shared by any target quota sharing user is less than the application quota, select multiple target quotas from the quotas shared by the multiple target quota sharing users, so that the sum of the multiple target quotas can be equal to the application quota.

[0288] It should be understood that the amount shared by the target amount sharing user will not be greater than the applied amount.

[0289] When the amount shared by a target quota sharing user is equal to the application amount, the amount shared by the target quota sharing user equal to the application amount can be directly received. When the amount shared by any target quota sharing user is less than the application amount, indicating that the remaining amount of all target quota sharing users is insufficient to share with the applicant, the terminal allocates the amount of multiple target quota sharing users, so that the target quota sharing users corresponding to the selected multiple target quotas can jointly complete the applicant's quota sharing application.

[0290] In an optional embodiment, the function providing module 208 is used to obtain a first target matching degree greater than a preset first matching degree threshold in the target social network graph, and provide a quota sharing reservation function to the contact corresponding to the first target matching degree.

[0291] In actual applications, when the applicant's credit limit is high and often unused, one or more contacts will invite him to share credit limit. The terminal can allow the applicant's contacts to make credit limit reservations by providing a credit limit sharing reservation function, so that the applicant can plan his credit limit in advance. Since the first contact corresponding to the first target matching degree in the target social network graph that is greater than the preset first matching degree threshold may have a very close relationship with the applicant, in order to be able to send credit limit sharing to the first contact, the terminal can provide a credit limit sharing reservation function to the contact corresponding to the first target matching degree, so that the first contact corresponding to the first target matching degree can send a credit limit sharing application to the applicant as soon as possible, so that the applicant can reserve credit limit for the first contact.

[0292] In an optional embodiment, the function providing module 208 is also used to obtain a second target matching degree in the target social network graph that is less than a preset second matching degree threshold, display a credit sharing authority option on the network node corresponding to the second target matching degree, receive a confirmed selection from the applicant in the credit sharing authority option, and refuse to receive a credit sharing application from the contact corresponding to the confirmed selection.

[0293] Wherein, the preset first matching degree threshold is greater than the preset second matching degree threshold.

[0294] The second contact corresponding to the second target matching degree in the target social network graph that is less than the preset second matching degree threshold is not very close to the applicant. In order to avoid receiving the credit sharing application from the second contact and disturbing the applicant, the credit sharing permission option is displayed on the network node corresponding to the second target matching degree in the target social network graph, so that the applicant can choose whether to reject the credit sharing application sent by the second contact. When the applicant clicks on the credit sharing permission option, the terminal receives the applicant's confirmation selection instruction in the credit sharing permission option, so that when receiving the credit sharing application of the contact corresponding to the confirmation selection, the terminal directly intercepts the credit sharing application of the contact corresponding to the confirmation selection to refuse to receive the credit sharing application of the contact corresponding to the confirmation selection.

[0295] In an optional implementation, the interest distribution module 209 is used to calculate additional interest and distribute benefits based on the risk score.

[0296] In order to solve the problem of overdraft and urgent need for credit turnover, the applicant should pay additional interest as the handling fee of the information service agency and the profit of the beneficiary, so that all three parties can benefit and achieve three goals at one stroke.

[0297] In an optional implementation, the interest distribution module 209 performs additional interest calculation and profit distribution according to the risk score, including:

[0298] Calculating the additional interest using an additional interest calculation model based on the risk score, the original credit limit of the applicant, and the application limit;

[0299] The distributed income of the target credit limit sharing users corresponding to the multiple target credit limits is calculated based on the additional interest and the second comprehensive credit score.

[0300] The additional interest calculation model can be the following formula: Y = (X1 + X2) / X3, where Y represents the additional interest, X1 represents the risk score, X2 represents the application amount, and X3 represents the original amount. It can be seen that the higher the risk score, the larger the application amount, the smaller the original amount, and the more additional interest; the lower the risk score, the smaller the application amount, the larger the original amount, and the less additional interest.

[0301] The terminal calculates the credit score sum of the second comprehensive credit scores of the target credit sharing users corresponding to the multiple target credits, and then calculates the ratio of the second comprehensive credit score of the target credit sharing user corresponding to each target credit to the credit score sum, and obtains the income of the corresponding target credit sharing user according to the product of the ratio and the additional interest. The higher the second comprehensive credit score, the larger the corresponding ratio, and the higher the income of the corresponding target credit sharing user; the lower the second comprehensive credit score, the smaller the corresponding ratio, and the lower the income of the corresponding target credit sharing user.

[0302] In this optional implementation, additional interest calculation and profit distribution are performed according to the risk score, which can increase the user's willingness to share the credit limit, motivate the user's trustworthy behavior, improve the user's credit score, and improve the credit service market.

[0303] The terminal may also share the user's earnings according to the product of the ratio and the preset ratio of the additional interest as the corresponding target amount.

[0304] In an optional embodiment, the score updating module 210 is used to obtain the credit behavior of the applicant with respect to the credit shared by the target credit sharing user; and update the first comprehensive credit score of the applicant and the second comprehensive credit score of the target credit sharing user according to the credit behavior.

[0305] After the target credit sharing user shares the credit with the applicant, it indicates that the applicant has successfully applied for the credit, and the terminal obtains the credit behavior of the applicant thereafter, which may include whether the books are returned on time or whether the loan is repaid on time. If the applicant returns the books or repays the loan on time, the credit behavior is considered to be trustworthy, and if the applicant fails to return the books or repay the loan on time, the credit behavior is considered to be untrustworthy.

[0306] The trustworthy behavior of the applicant also serves as the trustworthy behavior of the target credit sharing user, and the first comprehensive credit score of the applicant and the second comprehensive credit score of the target credit sharing user are updated respectively, and the social association between the applicant and the target credit sharing user is updated, thereby updating the matching degree between the applicant and the target credit sharing user, and then updating the social network map of the applicant. In this way, the target credit sharing user can guarantee the credit of the applicant, promote the trustworthy behavior of the applicant, and protect the interests of the credit service agency.

[0307] In an optional embodiment, the terminal can also store the credit sharing. When the applicant sends a second credit sharing application to the target credit sharing user again, the weight of the edge in the social network graph of the applicant is increased, thereby making the applicant's second credit sharing application more likely to succeed, reducing the computational complexity, and at the same time reducing the risk score.

[0308] See also Figure 4 FIG. 3 is a schematic diagram of the structure of a terminal provided in Embodiment 3 of the present invention. In a preferred embodiment of the present invention, the terminal 3 includes a memory 31, at least one processor 32, at least one communication bus 33 and a transceiver 34. The terminal may include a computer device.

[0309] Those skilled in the art should understand that Figure 4 The structure of the terminal shown does not constitute a limitation of the embodiments of the present invention, and may be either a bus structure or a star structure. The terminal 3 may also include more or less other hardware or software than shown in the figure, or a different arrangement of components.

[0310] In some embodiments, the terminal 3 is a device that can automatically perform numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes but is not limited to microprocessors, application-specific integrated circuits, programmable gate arrays, digital processors, and embedded devices. The terminal 3 may also include client devices, which include but are not limited to any electronic product that can interact with a client through a keyboard, mouse, remote control, touchpad, or voice-controlled device, such as a personal computer, tablet computer, smart phone, digital camera, etc.

[0311] It should be noted that the terminal 3 is only an example, and other existing or future electronic products that are suitable for the present invention should also be included in the protection scope of the present invention and included here by reference.

[0312] In some embodiments, the memory 31 stores a computer program, and when the computer program is executed by the at least one processor 32, all or part of the steps in the credit line sharing method based on big data are implemented. The memory 31 includes a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.

[0313] Furthermore, the computer-readable storage medium may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function, etc.; the data storage area may store data created according to the use of the blockchain node, etc.

[0314] In some embodiments, the at least one processor 32 is the control core (ControlUnit) of the terminal 3, and uses various interfaces and lines to connect various components of the entire terminal 3, and executes various functions and processes data of the terminal 3 by running or executing programs or modules stored in the memory 31, and calling data stored in the memory 31. For example, when the at least one processor 32 executes the computer program stored in the memory, it implements all or part of the steps of the credit line sharing method based on big data described in the embodiment of the present invention; or implements all or part of the functions of the credit line sharing device based on big data. The at least one processor 32 can be composed of an integrated circuit, for example, it can be composed of a single packaged integrated circuit, or it can be composed of multiple integrated circuits with the same function or different functions, including one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and various control chips.

[0315] In some embodiments, the at least one communication bus 33 is configured to implement connection and communication between the memory 31 and the at least one processor 32, etc.

[0316] Although not shown, the terminal 3 may also include a power source (such as a battery) for supplying power to various components. Preferably, the power source may be logically connected to the at least one processor 32 through a power management device, so that the power management device can manage charging, discharging, and power consumption management. The power source may also include one or more DC or AC power sources, recharging devices, power failure detection circuits, power converters or inverters, power status indicators, and other arbitrary components. The terminal 3 may also include a variety of sensors, Bluetooth modules, Wi-Fi modules, etc., which will not be described in detail here.

[0317] The above-mentioned integrated unit implemented in the form of a software function module can be stored in a computer-readable storage medium. The above-mentioned software function module is stored in a storage medium and includes a number of instructions for enabling a computer device (which can be a personal computer, terminal, or network device, etc.) or a processor to execute a part of the method described in each embodiment of the present invention.

[0318] In the several embodiments provided by the present invention, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only illustrative, for example, the division of the modules is only a logical function division, and there may be other division methods in actual implementation.

[0319] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical units, and may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0320] In addition, each functional module in each embodiment of the present invention may be integrated into one processing unit, each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of hardware plus software functional modules.

[0321] It is obvious to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential features of the present invention. Therefore, from any point of view, the embodiments should be regarded as exemplary and non-restrictive, and the scope of the present invention is defined by the appended claims rather than the above description, and it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims are included in the present invention. Any figure mark in the claims should not be regarded as limiting the claims involved. In addition, it is obvious that the word "including" does not exclude other units or, and the singular does not exclude the plural. Multiple units or devices stated in the specification may also be implemented by one unit or device through software or hardware. The words first, second, etc. are used to indicate names, and do not indicate any particular order.

[0322] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention.

Claims

1. A credit quota sharing method based on big data, characterized in that: The method comprises: In response to an applicant's quota application instruction, obtain the applicant's user data and application quota; Calculating a first comprehensive credit score based on the user data; Calculating a risk score based on the user data and the first comprehensive credit score using a long short-term memory network model; When it is determined that the risk score is less than or equal to a preset risk score threshold, constructing a social network graph of the applicant; Determine a plurality of target quota sharing users based on the social network graph; Sending a quota sharing application to the multiple target quota sharing users; The constructing of the social network graph of the applicant includes: Obtaining multiple contacts of the applicant; Calculating the social connection between the applicant and each contact; Constructing a social network graph by taking the applicant as the central node in the social network graph, taking the contacts as network nodes, and taking the social association degree as the weight of the edge between the central node and the corresponding network node; The calculating of the social connection between the applicant and each contact comprises: Calculate the social frequency and social time of the applicant and each contact; Extracting the social frequency and the social time to construct a social feature vector; Calculate the social relevance between the applicant and each contact based on the social feature vector using a social relevance calculation model; The determining of a plurality of target quota sharing users based on the social network graph comprises: Obtaining the remaining credit limit and second comprehensive credit score of the contact; constructing a contact feature vector based on the social association degree, the remaining credit limit and the second comprehensive credit score; Calculate the matching degree between the contact and the applicant based on the contact feature vector using a matching degree calculation model; According to the matching degree, the social network graph is updated to obtain a target social network graph; Search the target social network graph according to preset search rules to obtain multiple target network nodes; Determining the multiple target network nodes as multiple target quota sharing users; The preset search rule is to start from the central node and search for contacts corresponding to network nodes whose paths are no more than a preset number as the target quota sharing users of the applicant.

2. The credit line sharing method based on big data as claimed in claim 1, characterized in that: The user data includes basic data, credit limit usage data, historical credit data, third-party credit score data, and user social data. The calculation of the first comprehensive credit score based on the user data includes: Calculate user portrait based on the basic data; Predicting the user's spending capacity based on the credit usage data; Predicting the user's ability to fulfill a contract based on the historical credit data; The first comprehensive credit score is calculated based on the user portrait, the user's consumption ability, the user's contract performance ability, the third-party credit score data and the user's social data.

3. The credit line sharing method based on big data as claimed in claim 2, characterized in that: Calculating the user portrait based on the basic data includes: Acquire first data corresponding to the first data label and second data corresponding to the second data label in the basic data; Extracting a first feature vector of the first data and a second feature vector of the second data using a pre-trained model; A user portrait is calculated based on the first feature vector and the corresponding first weight, the second feature vector and the corresponding second weight.

4. The credit line sharing method based on big data as claimed in claim 2, characterized in that: The predicting of the user's consumption capacity based on the credit usage data includes: Calculating usage habits, usage frequency, credit limit utilization rate and credit limit overdraft rate based on the credit limit usage data; Constructing a user behavior preference vector according to the usage habits, the usage frequency, the credit limit usage rate and the credit limit overdraft rate; The trained consumption capacity prediction model is used to predict the user's consumption capacity based on the user behavior preference vector.

5. The credit line sharing method based on big data as claimed in claim 2, characterized in that: The predicting of the user's performance ability based on the historical credit data includes: Obtaining the fulfillment date and credit card debt in the historical credit data; Performing a first binning process on the performance date to obtain a binning date; Performing a second binning process on the credit card liabilities to obtain binned liabilities; Concatenating the binning date and the binning liability to obtain a user performance vector; The trained fulfillment capability prediction model is used to predict the user's fulfillment capability based on the user fulfillment vector.

6. The credit line sharing method based on big data as claimed in claim 2, characterized in that: The calculation process of the third-party credit score data includes: receiving credit score data of the applicant sent by a plurality of preset data service structures; Allocating scoring weights according to the sending time of the preset data service agency; The third-party credit score data is calculated based on the credit score data and the corresponding score weight.

7. The credit line sharing method based on big data as claimed in claim 1, characterized in that: After sending the quota sharing application to the plurality of target quota sharing users, the method further includes: Receiving the amount shared by the target amount sharing user; Determine whether the amount shared by any target amount sharing user is equal to the applied amount; When the amount shared by any target amount sharing user is equal to the applied amount, receiving the amount shared by any target amount sharing user; When the amount shared by any target amount sharing user is less than the applied amount, multiple target amounts are selected from the amounts shared by the multiple target amount sharing users so that the sum of the multiple target amounts can be equal to the applied amount.

8. The credit line sharing method based on big data as claimed in claim 1, characterized in that: The method further comprises: Obtain a first target matching degree in the target social network graph that is greater than a preset first matching degree threshold, and provide a credit sharing reservation function to the contact corresponding to the first target matching degree; or obtain a second target matching degree in the target social network graph that is less than a preset second matching degree threshold, display a credit sharing authority option on the network node corresponding to the second target matching degree, receive the applicant's confirmation selection in the credit sharing authority option, and refuse to receive the credit sharing application from the contact corresponding to the confirmation selection.

9. The credit line sharing method based on big data according to any one of claims 1 to 8, characterized in that: The method further comprises: Calculating additional interest and distributing benefits according to the risk score includes: using an additional interest calculation model to calculate the additional interest based on the risk score, the applicant's original credit limit and the applied credit limit; calculating the distributed benefits of the target credit limit sharing users corresponding to the multiple target credit limits according to the additional interest and the second comprehensive credit score.

10. The credit line sharing method based on big data as claimed in claim 9, characterized in that: The additional interest calculation model is the following formula: Y=(X1+X2) / X3, where Y represents the additional interest, X1 represents the risk score, X2 represents the application amount, and X3 represents the original amount.

11. The credit line sharing method based on big data according to claim 10, characterized in that: The method further comprises: Acquiring the credit behavior of the applicant with respect to the credit limit shared by the target credit limit sharing user; The first comprehensive credit score of the applicant and the second comprehensive credit score of the target credit limit sharing user are updated according to the credit behavior.

12. The credit line sharing method based on big data according to claim 10, characterized in that: The method further comprises: Obtaining account information of the applicant; Determining whether the remaining credit limit in the account information is a preset credit limit threshold; When it is determined that the remaining credit limit is the preset credit limit threshold, executing a response to the credit limit application instruction; When it is determined that the remaining credit limit is not the preset credit limit threshold, the response to the credit limit application instruction is rejected.

13. The credit line sharing method based on big data according to claim 10, characterized in that: The method further comprises: The target quota sharing user is identified as the applicant's guarantor.

14. A credit quota sharing device based on big data, characterized in that: The device comprises: An application response module, used to respond to an applicant's quota application instruction and obtain the applicant's user data and application quota; A first calculation module, configured to calculate a first comprehensive credit score based on the user data; A second calculation module, configured to calculate a risk score based on the user data and the first comprehensive credit score using a long short-term memory network model; A graph construction module is used to construct a social network graph of the applicant when it is determined that the risk score is less than or equal to a preset risk score threshold; the construction of the social network graph of the applicant includes: obtaining multiple contacts of the applicant; calculating the social association between the applicant and each contact; taking the applicant as the central node in the social network graph, the contact as the network node, and the social association as the weight of the edge between the central node and the corresponding network node to construct a social network graph; the calculation of the social association between the applicant and each contact includes: calculating the social frequency and social time of the applicant and each contact; extracting the social frequency and the social time to construct a social feature vector; using a social association calculation model to calculate the social association between the applicant and each contact based on the social feature vector; A target determination module is used to determine multiple target credit sharing users based on the social network graph; the determination of multiple target credit sharing users based on the social network graph includes: obtaining the remaining credit and the second comprehensive credit score of the contact; constructing a contact feature vector based on the social association, the remaining credit and the second comprehensive credit score; using a matching fit calculation model to calculate the matching fit between the contact and the applicant based on the contact feature vector; updating the social network graph according to the matching fit to obtain a target social network graph; searching the target social network graph according to a preset search rule to obtain multiple target network nodes; determining the multiple target network nodes as multiple target credit sharing users; The preset search rule is to start from the central node and search for contacts corresponding to network nodes with a path not exceeding a preset number as the target quota sharing users of the applicant; The sharing sending module is used to send a quota sharing application to the multiple target quota sharing users.

15. A terminal, characterized in that: The terminal includes a processor, and the processor is used to implement the credit line sharing method based on big data as described in any one of claims 1 to 14 when executing a computer program stored in a memory.

16. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the credit line sharing method based on big data as described in any one of claims 1 to 14 is implemented.

Images