Credit value evaluation method and device, equipment, storage medium and product
By combining user credit assessment information and product portrait information and using the trained credit assessment model to perform credit assessment, the problem of low credit assessment accuracy in the existing technology is solved, and more accurate credit value assessment and reasonable pre-credit limit allocation are achieved.
Patent Information
- Application Number
- CN202410302434.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-03-15
- Publication Date
- 2025-09-16
AI Technical Summary
Existing technologies do not consider users' personal emotional orientation towards products or services provided by credit merchants in credit value assessment, resulting in low accuracy of credit assessment.
By obtaining the credit assessment information of the user to be assessed and the contract product information of credit consumption, the credit assessment indicators are determined, and the trained target credit assessment model is used to combine the user's consumption sentiment information and product portrait information to perform credit assessment and improve the assessment accuracy.
It effectively improves the accuracy of credit assessment, avoids homogenization, and ensures that the assessment results are more in line with user consumption needs and merchant interests.
Smart Images
Figure CN120655404A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of data processing technology, and in particular to a credit value assessment method, device, equipment, storage medium and product. Background Art
[0002] With the development of mobile Internet, more and more financial institutions are providing consumer credit services to individual customers. The credit limit of credit services is allocated based on the user's credit score. In the current credit consumption process, the credit limit is directly determined by the financial institution based on the user's credit assessment results. The merchant providing the product is not involved. In addition, when evaluating the user's credit score, the user's personal emotional orientation towards the products or services provided by the credit merchant is not taken into account, which leads to a low accuracy in the final assessment of the user's credit score.
[0003] The above content is only used to assist in understanding the technical solution of the present invention and does not constitute an admission that the above content is prior art. Summary of the Invention
[0004] The main purpose of the present invention is to provide a credit value evaluation method, device, equipment, storage medium and product, aiming to solve the technical problem of low accuracy of credit value evaluation in the existing technology.
[0005] To achieve the above object, the present invention provides a credit evaluation method, which includes the following steps:
[0006] Obtain credit assessment information of the user to be assessed and contract product information for credit consumption;
[0007] Determining a credit assessment index based on the credit assessment information and the contract commodity information;
[0008] The current credit value of the user to be evaluated is evaluated according to the credit evaluation index based on a target credit evaluation model, wherein the target credit evaluation model is trained based on user consumption emotion information and product portrait information.
[0009] Optionally, before evaluating the current credit value of the user to be evaluated according to the credit evaluation index based on the target credit evaluation model, the method further includes:
[0010] Get the product order data of each merchant;
[0011] Generate product portrait information based on the product order data;
[0012] Determine the user information applicable to each product based on the product portrait information;
[0013] Mining user consumption sentiment information based on user information applicable to each product;
[0014] A target credit assessment model is trained based on the user consumption emotion information, the product portrait information, and the credit risk assessment model to be learned.
[0015] Optionally, the training of a target credit assessment model based on the user consumption emotion information, the product portrait information, and the credit risk assessment model to be learned includes:
[0016] Obtain historical credit information of each user;
[0017] determining a first credit assessment sample based on the historical credit information;
[0018] Obtaining a second credit assessment sample based on the user consumption emotion information and the product portrait information;
[0019] Evaluate the credit assessment initial value corresponding to the first credit assessment sample and the credit assessment standard value corresponding to the second credit assessment sample respectively using an initial credit assessment model;
[0020] The target credit assessment model is trained according to the credit assessment initial value, the credit assessment standard value and the credit risk assessment model to be learned.
[0021] Optionally, the training of a target credit assessment model based on the credit assessment initial value, the credit assessment standard value, and the credit risk assessment model to be learned includes:
[0022] Calculating the credit assessment difference between the initial credit assessment value and the standard credit assessment value;
[0023] Adjusting the parameters of the initial credit assessment model according to the assessment difference to obtain a first credit assessment model;
[0024] respectively evaluating the current credit assessment value corresponding to the third credit assessment sample and the current credit assessment standard value corresponding to the fourth credit assessment sample using the first credit assessment model;
[0025] Calculating a current assessment difference between the current credit assessment value and the current credit assessment standard value;
[0026] Adjusting the parameters of the first credit assessment model according to the current assessment difference to obtain a credit risk assessment model to be learned;
[0027] The target credit assessment model is trained based on the credit risk assessment model to be learned, the fitted user credit assessment information and contract product information sample set, and the preset assessment labels.
[0028] Optionally, the step of respectively evaluating the credit assessment initial value corresponding to the first credit assessment sample and the credit assessment standard value corresponding to the second credit assessment sample using the initial credit assessment model includes:
[0029] Cleaning the first credit assessment sample and the second credit assessment sample respectively;
[0030] Counting missing values of the cleaned first credit assessment sample and the cleaned second credit assessment sample respectively;
[0031] When the first credit assessment sample and the second credit assessment sample whose missing values are greater than the preset threshold satisfy a preset default relationship, filling the first credit assessment sample and the second credit assessment sample whose missing values are greater than the preset threshold;
[0032] Obtain a first target credit assessment sample and a second target credit assessment sample based on the credit assessment sample whose missing value is less than or equal to a preset threshold and the filled credit assessment sample;
[0033] processing the first target credit assessment sample and the second target credit assessment sample respectively to obtain a first derivative sample and a second derivative sample;
[0034] Performing normalization processing on the first derived sample and the second derived sample respectively;
[0035] Encoding the normalized first derivative sample and the second derivative sample respectively to obtain a first credit assessment feature and a second credit assessment feature;
[0036] The initial credit assessment value corresponding to the first credit assessment feature and the credit assessment standard value corresponding to the second credit assessment feature are respectively assessed using an initial credit assessment model.
[0037] Optionally, after evaluating the current credit value of the user to be evaluated according to the credit evaluation index based on the target credit evaluation model, the method further includes:
[0038] Determining a pre-credit limit for the user to be evaluated based on the current credit value;
[0039] Match target financial products based on the pre-approved credit limit.
[0040] In addition, to achieve the above-mentioned purpose, the present invention further provides a credit value evaluation device, which includes:
[0041] The acquisition module is used to obtain the credit assessment information of the user to be assessed and the contract product information of the credit consumption;
[0042] a determination module, configured to determine a credit assessment index based on the credit assessment information and the contract commodity information;
[0043] An evaluation module is used to evaluate the current credit value of the user to be evaluated according to the credit evaluation index based on a target credit evaluation model, wherein the target credit evaluation model is trained based on user consumption emotion information and product portrait information.
[0044] In addition, to achieve the above-mentioned purpose, the present invention also proposes a credit value assessment device, which includes: a memory, a processor, and a credit value assessment program stored in the memory and executable on the processor, wherein the credit value assessment program is configured to implement the credit value assessment method described above.
[0045] In addition, to achieve the above-mentioned purpose, the present invention further proposes a storage medium, on which a credit evaluation program is stored. When the credit evaluation program is executed by a processor, the credit evaluation method described above is implemented.
[0046] In addition, to achieve the above-mentioned object, the present invention further proposes a computer program product, which includes a credit value evaluation program. When the credit value evaluation program is executed by a processor, it implements the credit value evaluation method described above.
[0047] The credit value assessment method proposed in the present invention obtains the credit assessment information of the user to be assessed and the contract commodity information of the credit consumption; determines the credit assessment index based on the credit assessment information and the contract commodity information; evaluates the current credit value of the user to be assessed based on the credit assessment index based on the target credit assessment model, and evaluates the current credit value of the user to be assessed based on the credit assessment index based on the target credit assessment model, wherein the target credit assessment model is trained based on the user's consumption emotion information and commodity portrait information; through the above method, after obtaining the credit assessment information of the user to be assessed and the contract commodity information of the credit consumption, the credit assessment index is determined, and then the credit assessment index is evaluated using the trained target credit assessment model to obtain the current credit value of the user to be assessed, thereby effectively improving the accuracy of the credit value assessment and avoiding the occurrence of homogenization. BRIEF DESCRIPTION OF THE DRAWINGS
[0048] Figure 1 Schematic diagram of the structure of a credit value evaluation device in a hardware operating environment according to an embodiment of the present invention;
[0049] Figure 2 This is a flow chart of a first embodiment of the credit evaluation method of the present invention;
[0050] Figure 3 This is a flow chart of a second embodiment of the credit evaluation method of the present invention;
[0051] Figure 4 Schematic diagram of the functional modules of the first embodiment of the credit evaluation device of the present invention.
[0052] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION
[0053] It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.
[0054] Reference Figure 1 , Figure 1 This is a schematic diagram of the structure of a credit value assessment device in a hardware operating environment according to an embodiment of the present invention.
[0055] like Figure 1 As shown, the credit assessment device may include: a processor 1001, such as a central processing unit (CPU), a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. The communication bus 1002 is used to implement communication between these components. The user interface 1003 may include a display screen and an input unit such as a keyboard. Optionally, the user interface 1003 may also include a standard wired interface or a wireless interface. The network interface 1004 may optionally include a standard wired interface or a wireless interface (such as a wireless fidelity (Wi-Fi) interface). The memory 1005 may be a high-speed random access memory (RAM) or a stable non-volatile memory (NVM), such as a disk storage device. The memory 1005 may also be a storage device independent of the processor 1001.
[0056] Those skilled in the art will understand that Figure 1 The structure shown in the figure does not constitute a limitation to the credit evaluation device, and may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently.
[0057] like Figure 1 As shown, the memory 1005 as a storage medium may include an operating system, a network communication module, a user interface module, and a credit value evaluation program.
[0058] exist Figure 1In the credit evaluation device shown, the network interface 1004 is primarily used for data communication with a network integration platform workstation; the user interface 1003 is primarily used for data interaction with a user; the processor 1001 and memory 1005 in the credit evaluation device of the present invention can be provided within the credit evaluation device. The credit evaluation device invokes a credit evaluation program stored in the memory 1005 via the processor 1001 and executes the credit evaluation method provided in an embodiment of the present invention.
[0059] Based on the above hardware structure, an embodiment of the credit value evaluation method of the present invention is proposed.
[0060] Reference Figure 2 , Figure 2 Schematic diagram of the first embodiment of the credit evaluation method of the present invention.
[0061] In a first embodiment, the credit evaluation method includes the following steps:
[0062] Step S10: Obtain the credit assessment information of the user to be assessed and the contracted product information of the credit consumption.
[0063] It should be noted that the execution entity of this embodiment is a credit value assessment device, and it can also be other devices that can achieve the same or similar functions, such as a credit assessment platform, etc. This embodiment does not limit this. In this embodiment, the credit assessment platform is used as an example for explanation.
[0064] It should be understood that credit assessment information refers to information used to assess the credit value of the user to be assessed, and the credit assessment information includes but is not limited to basic information and historical borrowing information, among which the basic information includes but is not limited to asset information, income information, credit certification information, and academic certification information, and the historical borrowing information includes but is not limited to the loan amount, loan term, loan success date, loan type, number of loans, and loan repayment date.
[0065] Step S20: determining a credit evaluation index based on the credit evaluation information and the contract product information.
[0066] It can be understood that the credit assessment index refers to the index used to evaluate the user's credit value. After obtaining the credit assessment information of the user to be evaluated and the contract product information of the credit consumption respectively, the credit assessment information and the contract product information are fitted with features to obtain the credit assessment index.
[0067] Step S30: Evaluate the current credit value of the user to be evaluated according to the credit evaluation index based on a target credit evaluation model, wherein the target credit evaluation model is trained based on user consumption emotion information and product portrait information.
[0068] It should be understood that the target credit assessment model refers to a credit assessment model that is trained after incorporating user consumption emotional information and product portrait information. The target credit assessment model takes into account user consumption emotional information and product portrait information, making the assessment of user credit value more accurate and reasonable, meeting the interests of merchants, and thus balancing the credit relationship among financial institutions, merchants and users. It should be noted that before using the target credit assessment model to evaluate the credit value, in addition to inputting the credit assessment indicators, it is also necessary to input the consumption emotional information of the user to be evaluated when consuming a certain product and the portrait information of the product.
[0069] Furthermore, after step S30, the method further includes: determining a pre-credit limit of the user to be evaluated based on the current credit value; and matching a target financial product based on the pre-credit limit.
[0070] It can be understood that the pre-credit limit refers to the maximum credit limit for customer reference approved by the financial institution based on the credit value. The pre-credit limit can meet the shopping needs of credit consumption users and ensure the low risk of the financial institution. Generally speaking, the higher the user's current credit value, the higher the corresponding pre-credit limit. The target financial product refers to the financial product that matches the pre-credit limit. Since the credit value of each user evaluated by the target credit assessment model is different, the credit limit allocated to the user by the financial institution has productized embodiment to avoid the occurrence of homogeneity.
[0071] This embodiment obtains the credit assessment information of the user to be assessed and the contract commodity information of the credit consumption; determines the credit assessment index based on the credit assessment information and the contract commodity information; evaluates the current credit value of the user to be assessed based on the credit assessment index based on the target credit assessment model, and evaluates the current credit value of the user to be assessed based on the credit assessment index based on the target credit assessment model, wherein the target credit assessment model is trained based on the user's consumption emotion information and commodity portrait information; through the above method, after obtaining the credit assessment information of the user to be assessed and the contract commodity information of the credit consumption, the credit assessment index is determined, and then the credit assessment index is evaluated using the trained target credit assessment model to obtain the current credit value of the user to be assessed, thereby effectively improving the accuracy of the assessed credit value and avoiding the occurrence of homogenization.
[0072] In one embodiment, if Figure 3 As shown, a second embodiment of the credit value evaluation method of the present invention is proposed based on the first embodiment. Before step S30, the method further includes:
[0073] Step S201: Obtain commodity order data of each merchant.
[0074] It should be understood that the product order data can be the order data of all categories of products for each merchant that has opened credit consumption, or it can be the product order data of each merchant that has added credit consumption on different platforms. This embodiment does not limit this. The product order data includes but is not limited to product name, product type, sales platform, user age, user gender, number of user consumption times, user consumption amount, user payment method, and user payment data.
[0075] Step S202: Generate product image information based on the product order data.
[0076] It is understandable that after obtaining the product order data of each merchant, product portrait information is generated through a portrait generation tool. The product portrait information includes but is not limited to product appearance attributes, supply chain attributes, positioning attributes, category attributes, application attributes, functional attributes, sales attributes and financial attributes, etc.
[0077] It should be understood that for credit payment, on the one hand, it is necessary to ensure that the credit line granted by the financial institution to the user brings the least credit risk to the financial institution, and on the other hand, it is also necessary to ensure that the pre-credit line allocated to the user can most reasonably meet the user's consumption needs, that is, to bring reasonable income to the merchants bound by the credit consumption. However, the pre-credit line in the existing technology is only determined based on the user's credit assessment information, and does not take into account the merchants providing services or products and the user's consumption willingness.
[0078] Step S203: Determine the user information applicable to each product based on the product portrait information.
[0079] It should be understood that different products have different applicable user groups. Therefore, after generating product portrait information, the applicable user information for each product is determined based on the product portrait information.
[0080] Step S204: mining user consumption emotion information based on the user information applicable to each commodity.
[0081] It is understandable that user consumption emotional information represents the user's personal emotional tendency to consume goods. The more inclined a user is to consume a certain product, the stronger the consumption intention for the product is. That is, the user information applicable to each product can be used to mine user consumption emotional information.
[0082] Step S205: training a target credit assessment model based on the user consumption emotion information, the product portrait information, and the credit risk assessment model to be learned.
[0083] It should be understood that the credit risk assessment model to be learned refers to a credit risk assessment model that has been updated multiple times with the assessment difference. After the user's consumption emotional information and product portrait information are combined with the preset assessment labels, the target credit assessment model is trained together.
[0084] Furthermore, step S205 includes: obtaining historical credit information of each user; determining a first credit assessment sample based on the historical credit information; obtaining a second credit assessment sample based on the user consumption emotion information and the product portrait information; respectively evaluating the credit assessment initial value corresponding to the first credit assessment sample and the credit assessment standard value corresponding to the second credit assessment sample through the initial credit assessment model; training the target credit assessment model based on the credit assessment initial value, the credit assessment standard value and the credit risk assessment model to be learned.
[0085] It can be understood that historical credit information refers to the relevant information of users' credit at historical moments. The historical credit information includes but is not limited to the basic information and historical borrowing information of each user at historical moments. The credit assessment platform then determines the first credit assessment sample based on the historical credit information. The credit assessment platform can be a financial institution or a third-party assessment agency independent of the financial institution. The second credit assessment sample refers to a credit assessment sample that has the willingness to consume a certain product. The second credit assessment sample can realize the constraints on the model training, so that the credit value output by the trained credit assessment model can be able to maximize the integration of user consumption emotional information. The credit assessment standard value refers to the dynamic label data used to complete the training of the initial credit assessment model.
[0086] It should be understood that after determining the first credit assessment sample and the second credit assessment sample, the initial credit assessment model is used to evaluate the credit assessment initial value and the credit assessment initial value respectively. When there are multiple initial credit assessment models, the credit assessment initial value can be a comprehensive value of multiple credit assessment initial values. The initial credit assessment model is deployed on the credit assessment platform and can be at least one of SVC, GBDT, RF, AdaBoost, XGBoost, LightGBM, KNN and LR. In addition, in this embodiment, the label value used for supervision model training is not predetermined, but is dynamically generated by the initial credit assessment model to be trained. The user's consumption sentiment information can be introduced into the training process of the credit risk assessment model through dynamic generation, thereby improving the accuracy of credit value evaluation using the credit assessment model.
[0087] Furthermore, the training of the target credit assessment model based on the credit assessment initial value, the credit assessment standard value and the credit risk assessment model to be learned includes: calculating the evaluation difference between the credit assessment initial value and the credit assessment standard value; adjusting the parameters of the initial credit assessment model according to the evaluation difference to obtain a first credit assessment model; respectively evaluating the current credit assessment value corresponding to the third credit assessment sample and the current credit assessment standard value corresponding to the fourth credit assessment sample through the first credit assessment model; calculating the current evaluation difference between the current credit assessment value and the current credit assessment standard value; adjusting the parameters of the first credit assessment model according to the current evaluation difference to obtain the credit risk assessment model to be learned; and training the target credit assessment model based on the credit risk assessment model to be learned, the fitted user credit assessment information and contract product information sample set, and the preset evaluation labels.
[0088] It should be understood that after obtaining the initial credit assessment value and the credit assessment standard value, the difference operation is performed between the two, and it is determined whether the assessment difference is greater than the preset threshold. If so, the gradient descent algorithm is used to adjust the parameters of the initial credit assessment model. The third credit assessment sample and the fourth credit assessment sample are different from the first credit assessment sample and the second credit assessment sample only in batches, and all belong to credit assessment samples. Then, the current assessment difference between the current credit assessment value and the current credit assessment standard value is calculated again. When the current assessment difference is still greater than the preset threshold, the parameter adjustment is continued, and the credit assessment sample is input again for learning until the final assessment difference is minimized (constant and infinitely close to the preset threshold, which can be 0), and the model is terminated. Model training is performed, and the final updated model is used as the credit risk assessment model to be learned. It should be noted that the credit assessment standard value each time is re-determined based on the output of the model after parameter adjustment, so that the model can fit the characteristic attributes of users with strong consumption intentions, and ensure that the credit value evaluated by the trained model is comprehensive of the multi-dimensional factors of risk and user consumption emotional information, so that the pre-credit limit finally allocated to the user can meet the shopping needs of credit consumption users and ensure the low-risk requirements of financial institutions. The preset assessment label can be a label determined according to the historical experience library to improve the accuracy of the training target credit assessment model. The preset assessment label corresponds to the fitted user credit assessment information and contract product information sample set.
[0089] Furthermore, the initial credit assessment model is used to respectively evaluate the credit assessment initial value corresponding to the first credit assessment sample and the credit assessment standard value corresponding to the second credit assessment sample, including: cleaning the first credit assessment sample and the second credit assessment sample respectively; counting the missing values of the cleaned first credit assessment sample and the cleaned second credit assessment sample respectively; when the first credit assessment sample and the second credit assessment sample whose missing values are greater than a preset threshold satisfy a preset default relationship, filling the first credit assessment sample and the second credit assessment sample whose missing values are greater than the preset threshold; obtaining a first target credit assessment sample and a second target credit assessment sample based on the credit assessment samples whose missing values are less than or equal to the preset threshold and the filled credit assessment samples; processing the first target credit assessment sample and the second target credit assessment sample respectively to obtain a first derivative sample and a second derivative sample; normalizing the first derivative sample and the second derivative sample respectively; encoding the normalized first derivative sample and the second derivative sample respectively to obtain a first credit assessment feature and a second credit assessment feature; and respectively evaluating the credit assessment initial value corresponding to the first credit assessment feature and the credit assessment standard value corresponding to the second credit assessment feature through the initial credit assessment model.
[0090] It is understandable that before evaluating the credit assessment value corresponding to the credit assessment sample through the initial credit assessment model, the credit assessment sample needs to be processed, specifically cleaning and preprocessing, that is, when the first credit assessment sample and the second credit assessment sample with missing values greater than the preset threshold meet the preset default relationship, "MISSING" is used to fill the first credit assessment sample and the second credit assessment sample with missing values greater than the preset threshold. After the cleaning is completed, the first target credit assessment sample and the second target credit assessment sample are processed respectively to generate the first derivative sample and the second derivative sample, and the processing method is different for different types of derivative samples. For example, for date-type derivative samples, the original value is replaced by the difference between the date and the earliest date of the derivative sample. For derivative samples with more categories, in order to reduce the amount of calculation and the number of categories, the first derivative sample and the second derivative sample are normalized and Onehot encoded respectively to obtain the credit assessment initial value corresponding to the first credit assessment feature and the credit assessment standard value corresponding to the second credit assessment feature.
[0091] This embodiment obtains the commodity order data of each merchant; generates commodity portrait information based on the commodity order data; determines the user information applicable to each commodity based on the commodity portrait information; mines the user consumption emotion information based on the user information applicable to each commodity; trains the target credit assessment model based on the user consumption emotion information, the commodity portrait information and the credit risk assessment model to be learned; through the above method, the user information applicable to each commodity is determined by using the commodity order data of each merchant, the user consumption emotion information is reversely mined by using the user information applicable to each commodity, and then the target credit assessment model is comprehensively trained in combination with the commodity portrait information and the credit risk assessment model to be learned, thereby effectively improving the accuracy of the trained target credit assessment model.
[0092] In addition, an embodiment of the present invention further provides a storage medium, on which a credit evaluation program is stored. When the credit evaluation program is executed by a processor, the steps of the credit evaluation method described above are implemented.
[0093] Since the storage medium adopts all the technical solutions of all the above embodiments, it has at least all the beneficial effects brought by the technical solutions of the above embodiments, which will not be described one by one here.
[0094] In addition, refer to Figure 4 The embodiment of the present invention further provides a credit value evaluation device, the credit value evaluation device comprising:
[0095] The acquisition module 10 is used to acquire the credit evaluation information of the user to be evaluated and the contract product information of the credit consumption.
[0096] The determination module 20 is configured to determine a credit evaluation index based on the credit evaluation information and the contract product information.
[0097] The evaluation module 30 is used to evaluate the current credit value of the user to be evaluated according to the credit evaluation index based on the target credit evaluation model, wherein the target credit evaluation model is trained based on user consumption emotion information and product portrait information.
[0098] This embodiment obtains the credit assessment information of the user to be assessed and the contract commodity information of the credit consumption; determines the credit assessment index based on the credit assessment information and the contract commodity information; evaluates the current credit value of the user to be assessed based on the credit assessment index based on the target credit assessment model, and evaluates the current credit value of the user to be assessed based on the credit assessment index based on the target credit assessment model, wherein the target credit assessment model is trained based on the user's consumption emotion information and commodity portrait information; through the above method, after obtaining the credit assessment information of the user to be assessed and the contract commodity information of the credit consumption, the credit assessment index is determined, and then the credit assessment index is evaluated using the trained target credit assessment model to obtain the current credit value of the user to be assessed, thereby effectively improving the accuracy of the assessed credit value and avoiding the occurrence of homogenization.
[0099] It should be noted that the workflow described above is merely illustrative and does not limit the scope of protection of the present invention. In practical applications, technicians in this field can select part or all of it according to actual needs to achieve the purpose of the embodiment scheme, and no limitation is made here.
[0100] In addition, for technical details not fully described in this embodiment, reference can be made to the credit value evaluation method provided in any embodiment of the present invention, and will not be repeated here.
[0101] In one embodiment, the evaluation module 30 is also used to obtain product order data of each merchant; generate product portrait information based on the product order data; determine the user information applicable to each product based on the product portrait information; mine user consumption sentiment information based on the user information applicable to each product; and train a target credit assessment model based on the user consumption sentiment information, the product portrait information, and the credit risk assessment model to be learned.
[0102] In one embodiment, the evaluation module 30 is further used to obtain historical credit information of each user; determine a first credit evaluation sample based on the historical credit information; obtain a second credit evaluation sample based on the user consumption emotion information and the product portrait information; evaluate the credit evaluation initial value corresponding to the first credit evaluation sample and the credit evaluation standard value corresponding to the second credit evaluation sample through the initial credit evaluation model; and train the target credit evaluation model based on the credit evaluation initial value, the credit evaluation standard value, and the credit risk assessment model to be learned.
[0103] In one embodiment, the evaluation module 30 is further used to calculate the evaluation difference between the initial credit evaluation value and the credit evaluation standard value; adjust the parameters of the initial credit evaluation model according to the evaluation difference to obtain a first credit evaluation model; respectively evaluate the current credit evaluation value corresponding to the third credit evaluation sample and the current credit evaluation standard value corresponding to the fourth credit evaluation sample through the first credit evaluation model; calculate the current evaluation difference between the current credit evaluation value and the current credit evaluation standard value; adjust the parameters of the first credit evaluation model according to the current evaluation difference to obtain a credit risk assessment model to be learned; and train the target credit assessment model according to the credit risk assessment model to be learned, the fitted user credit evaluation information and contract product information sample set, and the preset evaluation label.
[0104] In one embodiment, the evaluation module 30 is further used to clean the first credit assessment sample and the second credit assessment sample respectively; count the missing values of the cleaned first credit assessment sample and the cleaned second credit assessment sample respectively; fill the first credit assessment sample and the second credit assessment sample with missing values greater than the preset threshold when the first credit assessment sample and the second credit assessment sample with missing values greater than the preset threshold meet the preset default relationship; obtain the first target credit assessment sample and the second target credit assessment sample based on the credit assessment samples with missing values less than or equal to the preset threshold and the filled credit assessment samples; process the first target credit assessment sample and the second target credit assessment sample respectively to obtain the first derivative sample and the second derivative sample; normalize the first derivative sample and the second derivative sample respectively; encode the normalized first derivative sample and the second derivative sample respectively to obtain the first credit assessment feature and the second credit assessment feature; and evaluate the credit assessment initial value corresponding to the first credit assessment feature and the credit assessment standard value corresponding to the second credit assessment feature respectively through the initial credit assessment model.
[0105] In one embodiment, the evaluation module 30 is further configured to determine a pre-credit limit of the user to be evaluated based on the current credit value; and match a target financial product based on the pre-credit limit.
[0106] Other embodiments or implementation methods of the credit evaluation device of the present invention can refer to the above-mentioned method embodiments and will not be described in detail here.
[0107] It should be understood that, although the various steps in the flowchart in the embodiment of the present application are shown in sequence according to the indication of the arrows, these steps are not necessarily performed in sequence in the order indicated by the arrows. Unless clearly stated herein, the execution of these steps is not strictly limited in order, and they can be performed in other orders. Moreover, at least a portion of the steps in the figure may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily performed at the same time, but can be performed at different times, and their execution order is not necessarily performed in sequence, but can be performed in turn or alternately with at least a portion of other steps or sub-steps or stages of other steps.
[0108] In addition, it should be noted that, in this document, the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, article, or system comprising a series of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or system. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or system comprising the element.
[0109] The serial numbers of the above embodiments of the present invention are for description only and do not represent the advantages or disadvantages of the embodiments.
[0110] Through the description of the above embodiments, those skilled in the art can clearly understand that the above embodiment methods can be implemented by means of software plus the necessary general hardware platform, or of course by hardware, but in many cases the former is a better embodiment. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as a read-only memory (ROM) / RAM, a magnetic disk, or an optical disk), and includes a number of instructions for enabling a terminal device (which can be a mobile phone, a computer, an integrated platform workstation, or a network device, etc.) to execute the methods described in each embodiment of the present invention.
[0111] The above are only preferred embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention description and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.
Claims
1. A credit value evaluation method, characterized in that: The credit value evaluation method comprises the following steps: Obtain credit assessment information of the user to be assessed and contract product information for credit consumption; Determining a credit assessment index based on the credit assessment information and the contract commodity information; The current credit value of the user to be evaluated is evaluated according to the credit evaluation index based on a target credit evaluation model, wherein the target credit evaluation model is trained based on user consumption emotion information and product portrait information.
2. The credit evaluation method according to claim 1, wherein: Before evaluating the current credit value of the user to be evaluated according to the credit evaluation index based on the target credit evaluation model, the method further includes: Get the product order data of each merchant; Generate product portrait information based on the product order data; Determine the user information applicable to each product based on the product portrait information; Mining user consumption sentiment information based on user information applicable to each product; A target credit assessment model is trained based on the user consumption emotion information, the product portrait information, and the credit risk assessment model to be learned.
3. The credit evaluation method according to claim 2, wherein: The training of the target credit assessment model based on the user consumption emotion information, the product portrait information, and the credit risk assessment model to be learned includes: Obtain historical credit information of each user; determining a first credit assessment sample based on the historical credit information; Obtaining a second credit assessment sample based on the user consumption emotion information and the product portrait information; Evaluate the credit assessment initial value corresponding to the first credit assessment sample and the credit assessment standard value corresponding to the second credit assessment sample respectively using an initial credit assessment model; The target credit assessment model is trained according to the credit assessment initial value, the credit assessment standard value and the credit risk assessment model to be learned.
4. The credit evaluation method according to claim 3, wherein: The training of the target credit assessment model according to the credit assessment initial value, the credit assessment standard value, and the credit risk assessment model to be learned includes: Calculating the credit assessment difference between the initial credit assessment value and the standard credit assessment value; Adjusting the parameters of the initial credit assessment model according to the assessment difference to obtain a first credit assessment model; respectively evaluating the current credit assessment value corresponding to the third credit assessment sample and the current credit assessment standard value corresponding to the fourth credit assessment sample using the first credit assessment model; Calculating a current assessment difference between the current credit assessment value and the current credit assessment standard value; Adjusting the parameters of the first credit assessment model according to the current assessment difference to obtain a credit risk assessment model to be learned; The target credit assessment model is trained based on the credit risk assessment model to be learned, the fitted user credit assessment information and contract product information sample set, and the preset assessment labels.
5. The credit evaluation method according to claim 3, wherein: The step of respectively evaluating the credit assessment initial value corresponding to the first credit assessment sample and the credit assessment standard value corresponding to the second credit assessment sample using the initial credit assessment model includes: Cleaning the first credit assessment sample and the second credit assessment sample respectively; Counting missing values of the cleaned first credit assessment sample and the cleaned second credit assessment sample respectively; When the first credit assessment sample and the second credit assessment sample whose missing values are greater than the preset threshold satisfy a preset default relationship, filling the first credit assessment sample and the second credit assessment sample whose missing values are greater than the preset threshold; Obtain a first target credit assessment sample and a second target credit assessment sample based on the credit assessment sample whose missing value is less than or equal to a preset threshold and the filled credit assessment sample; processing the first target credit assessment sample and the second target credit assessment sample respectively to obtain a first derivative sample and a second derivative sample; Performing normalization processing on the first derived sample and the second derived sample respectively; Encoding the normalized first derivative sample and the second derivative sample respectively to obtain a first credit assessment feature and a second credit assessment feature; The initial credit assessment value corresponding to the first credit assessment feature and the credit assessment standard value corresponding to the second credit assessment feature are respectively assessed using an initial credit assessment model.
6. The credit evaluation method according to any one of claims 1 to 5, characterized in that: After evaluating the current credit value of the user to be evaluated according to the credit evaluation index based on the target credit evaluation model, the method further includes: Determining a pre-credit limit for the user to be evaluated based on the current credit value; Match target financial products based on the pre-approved credit limit.
7. A credit value evaluation device, characterized in that: The credit value evaluation device includes: The acquisition module is used to obtain the credit assessment information of the user to be assessed and the contract product information of the credit consumption; a determination module, configured to determine a credit assessment index based on the credit assessment information and the contract commodity information; An evaluation module is used to evaluate the current credit value of the user to be evaluated according to the credit evaluation index based on a target credit evaluation model, wherein the target credit evaluation model is trained based on user consumption emotion information and product portrait information.
8. A credit value evaluation device, characterized in that: The credit evaluation device includes a memory, a processor, and a credit evaluation program stored in the memory and executable on the processor. The credit evaluation program is configured to implement the credit evaluation method according to any one of claims 1 to 6.
9. A storage medium, characterized in that: The storage medium stores a credit evaluation program, which, when executed by a processor, implements the credit evaluation method according to any one of claims 1 to 6.
10. A computer program product, characterized in that The computer program product includes a credit evaluation program, and when the credit evaluation program is executed by a processor, the credit evaluation method according to any one of claims 1 to 6 is implemented.