Early warning method and device for user risk, equipment and storage medium
By generating training data sets, building risk identification models and real-time risk assessment methods, the shortcomings of the risk warning methods in the steel industry in terms of business feature analysis, model update and risk quantification are solved, and the accuracy and timeliness of risk warning are improved.
Patent Information
- Application Number
- CN202411884557.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-19
- Publication Date
- 2025-04-18
AI Technical Summary
The existing risk warning methods in the steel industry are insufficient in business characteristics analysis, model updates and risk quantification, resulting in insufficient accuracy and timeliness of early warning results.
By generating training data sets from multiple data sources, extracting data features to form a feature library, building a risk identification model layered, and conducting risk assessment based on real-time data to achieve accurate early warning of user risks.
It improves the accuracy and timeliness of risk warning, can more accurately identify potential risks, and timely adapt to market changes through dynamic model update mechanisms, real-time quantitative assessment is achieved.
Smart Images
Figure CN119849922A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of data processing technology, and in particular to a user risk early warning method, device, equipment and storage medium. Background Art
[0002] At present, risk early warning technologies in various industries have made significant progress. In the steel industry, existing risk early warning technologies mainly focus on the analysis of financial statements, such as ratio analysis and trend analysis, to identify potential financial risks.
[0003] The inventors of the present application have found that the limitations of existing risk early warning methods in the steel industry include the following:
[0004] (1) Insufficient analysis of business characteristics: Although existing methods have incorporated business characteristics, they are still lacking in in-depth analysis of business characteristics and have failed to fully explore potential risk points in business processes.
[0005] (2) Model update lag: As the market environment changes, existing methods mostly rely on historical data and fail to fully consider the potential default risks of the steel industry and upstream and downstream companies. There is a certain lag in model updates, and the early warning model needs to be continuously updated to maintain its effectiveness.
[0006] (3) Risk quantification is insufficient: Existing methods usually focus on risk identification and early warning, but lack real-time or quasi-real-time quantitative assessment and are unable to quantify potential risks based on changes in the steel industry market.
[0007] It can be seen that the existing risk warning methods in the steel industry have significantly limited the accuracy and timeliness of the warning results. Therefore, how to quantify and warn users of potential risks based on the steel industry transaction process and combine financial analysis to achieve the goal of risk monitoring and warning has become an urgent problem to be solved in the steel industry. Summary of the invention
[0008] According to one aspect of the present application, a user risk early warning method is provided, comprising: generating a training data set based on user historical data from multiple data sources; extracting data features corresponding to the training data set to form a feature library; based on the feature library, hierarchically constructing a risk identification model; based on the risk identification model, performing risk assessment on user real-time data.
[0009] According to some embodiments, generating a training data set based on user historical data from multiple data sources includes: acquiring user historical data; preprocessing the user historical data; and generating training data sets corresponding to the multiple data sources based on the preprocessed user historical data.
[0010] According to some embodiments, data features corresponding to a training data set are extracted to form a feature library, including: setting a plurality of candidate feature libraries; determining a risk target variable in the training data set; determining the data features corresponding to the training data set through a preset feature selection model based on the risk target variable; and determining a feature library from a plurality of candidate feature libraries based on the data features corresponding to the training data set.
[0011] According to some embodiments, a risk identification model is hierarchically constructed based on a feature library, including: dividing user groups to determine user stratification; based on the user stratification, a risk identification model is hierarchically constructed based on the feature library to output the user's risk occurrence probability; the risk identification model is trained through a training data set; and the trained risk identification model is optimized and verified.
[0012] According to some embodiments, an expected risk loss calculation formula is set according to the probability of risk occurrence.
[0013] According to some embodiments, a formula for calculating expected risk losses is set based on the probability of risk occurrence, including: obtaining a default loss rate through user historical data; obtaining the actual amount corresponding to the user risk when it occurs; and setting a formula for calculating expected risk losses based on the probability of risk occurrence, the default loss rate and the actual amount corresponding to the user risk when it occurs.
[0014] According to some embodiments, risk assessment of user real-time data is performed based on a risk identification model, including: inputting user real-time data into the risk identification model; obtaining the user's risk occurrence probability and expected risk loss through the risk identification model and the expected risk loss calculation formula; based on the risk occurrence probability and the expected risk loss, generating warning information according to a preset risk warning threshold.
[0015] According to one aspect of the present application, there is provided an early warning device for user risks, comprising: a data acquisition module for generating a training data set based on user historical data from multiple data sources; a feature extraction module for extracting data features corresponding to the training data set to form a feature library; a model construction module for hierarchically constructing a risk identification model based on the feature library; and a risk assessment module for performing risk assessment of user real-time data based on the risk identification model.
[0016] According to one aspect of the present application, an electronic device is provided, comprising: one or more processors; a storage device for storing one or more programs; when the one or more programs are executed by the one or more processors, the one or more processors implement the aforementioned method.
[0017] According to one aspect of the present application, a computer-readable storage medium is provided, on which a computer program is stored, and when the program is executed by a processor, the aforementioned method is implemented.
[0018] According to the embodiments of the present application, starting from the steel trading business process, combined with business characteristics and financial analysis, potential risks can be more accurately identified, the accuracy of risk warnings can be improved, and the dynamic model update mechanism can ensure that the warning model can adapt to market changes in a timely manner, thereby improving the timeliness of risk warnings. In addition, real-time quantitative assessments can be performed based on market changes, risk estimates of upstream and downstream enterprises, etc., to improve the observability of risk warnings.
[0019] It should be understood that the foregoing general description and the following detailed description are exemplary only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS
[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application.
[0021] Figure 1 A flow chart showing a method for early warning of user risk according to an exemplary embodiment of the present application.
[0022] Figure 2 A block diagram of a user risk early warning device according to an exemplary embodiment of the present application is shown.
[0023] Figure 3 A block diagram of an electronic device according to an exemplary embodiment of the present application is shown. DETAILED DESCRIPTION
[0024] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in many forms and should not be construed as limited to the embodiments set forth herein; rather, these embodiments are provided so that this application will be comprehensive and complete and will fully convey the concepts of the example embodiments to those skilled in the art. The same reference numerals in the figures represent the same or similar parts, and thus their repeated description will be omitted.
[0025] The described features, structures or characteristics may be combined in one or more embodiments in any suitable manner. In the following description, many specific details are provided to provide a full understanding of the embodiments of the present application. However, those skilled in the art will appreciate that the technical scheme of the present application may be put into practice without one or more of these specific details, or other modes, components, materials, devices or operations may be adopted. In these cases, known structures, methods, devices, realizations, materials or operations will not be shown or described in detail.
[0026] The flowcharts shown in the accompanying drawings are only exemplary and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps can be decomposed, and some operations / steps can be combined or partially combined, so the actual execution order may change according to actual conditions.
[0027] The terms "first", "second", etc. in the specification and claims of this application and the above-mentioned drawings are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally includes steps or units that are not listed, or optionally includes other steps or units inherent to these processes, methods, products or devices.
[0028] The present application provides a user risk early warning method, device, equipment and storage medium, which can improve the accuracy, timeliness and observability of risk early warning.
[0029] A user risk warning method, apparatus, device and storage medium according to an embodiment of the present application will be described in detail below with reference to the accompanying drawings.
[0030] Figure 1 A flow chart showing a method for early warning of user risk according to an exemplary embodiment of the present application.
[0031] like Figure 1 As shown, in step S100, a training data set is generated based on user historical data from multiple data sources.
[0032] For example, in step S100, the early warning device preprocesses the user historical data obtained from multiple data sources and generates a corresponding training data set.
[0033] The early warning device first obtains user historical data from multiple data sources.
[0034] According to some embodiments, the early warning device can obtain user historical data from multiple data sources by timed crawling or batch importing according to the results of the previous data exploration and analysis and the achievement of goals. The multiple data sources include user basic information, business system transaction data, financial data, and external market data.
[0035] The early warning device pre-processes the acquired user historical data.
[0036] According to some embodiments, the early warning device aligns the user historical data of multiple data sources by date and user dimension to ensure the consistency and integrity of the data. In addition, the early warning device also cleans the user historical data to handle missing values and abnormal values in the data.
[0037] According to some embodiments, the early warning device can select corresponding preprocessing methods according to different indicator data. For example, for continuous data, a preprocessing method of no processing can be selected, or a preprocessing method of data standardization, binning, etc. can be selected. For categorical data, a preprocessing method of encoding conversion can be selected.
[0038] After preprocessing the user history data, the early warning device generates a training data set corresponding to multiple data sources based on the preprocessed user history data.
[0039] In step S200, data features corresponding to the training data set are extracted to form a feature library.
[0040] For example, in step S200, the early warning device extracts data features corresponding to the training data set through a preset feature selection model, and determines a feature library based on this.
[0041] The early warning device sets up multiple candidate feature libraries through preliminary steel transaction business processes and financial analysis.
[0042] According to some embodiments, the plurality of candidate feature libraries include a candidate feature library corresponding to user basic information, a candidate feature library corresponding to business features, a candidate feature library corresponding to financial features, and a candidate feature library corresponding to external market features.
[0043] Among them, the candidate feature library corresponding to the user's basic information includes user equity relationship, user type (such as upstream supplier or trader or others), whether it is a listed company, whether a long-term cooperation agreement has been signed, registered scale and paid-in capital, etc.
[0044] The candidate feature library corresponding to the business features includes transaction conditions (such as the total business volume in the past year, the average monthly business volume, the invoice issuance cycle, the business sales cycle, and the number of days since the most recent transaction, etc.), business abnormality indicators (such as whether there is a registration abnormality, whether there has been an administrative warning in the past year, whether there have been industrial and commercial changes, whether there have been judicial warnings, whether there are negative information and related relationships, etc.).
[0045] The candidate feature library corresponding to financial features includes basic operating aspects (such as gross profit margin, operating income growth rate and gross profit growth rate, etc.), management aspects (such as inventory turnover days, accounts receivable turnover days and prepaid accounts turnover days, etc.), financial aspects (such as current ratio, debt-to-asset ratio and interest-bearing debt ratio, etc.), performance aspects (such as net profit growth rate, total profit growth rate, net profit attributable to shareholders and return on equity, etc.), and performance (such as considering the number of defaults, credit overdue days and payment timeliness, etc.).
[0046] The candidate feature library corresponding to external market characteristics includes price fluctuations of upstream raw materials for steel, annual steel output, capacity utilization, steel plant inventory, social inventory and downstream steel industry characteristics.
[0047] The early warning device determines the risk target variables in the training data set.
[0048] According to some embodiments, the early warning device first obtains the manually annotated data generated when the training data set is manually annotated according to the actual situation. Then, based on the manually annotated data of the training data set, the early warning device determines the default / overdue risk of the users in the training data set as the risk target variable. Among them, users who have defaulted / overdue are marked as 1, and users who have not defaulted / overdue are marked as 0 (i.e., normal users).
[0049] According to the determined risk target variables, the early warning device determines the data features corresponding to the training data set through a preset feature selection model.
[0050] According to some embodiments, the early warning device may first extract data features corresponding to the training data set and form a feature set. Then, the early warning device gradually eliminates unimportant features in the feature set through recursive feature elimination to optimize the feature set.
[0051] Furthermore, the early warning device obtains a preset feature selection model that has been trained, and automatically screens important features in the optimized feature set through the preset feature selection model.
[0052] The early warning device obtains a preset correlation threshold, and uses it to continue to screen data features related to the risk target variable from the important features obtained through the preset feature selection model.
[0053] The early warning device uses the screened data features related to the risk target variable as the final data features corresponding to the training data set, and forms a feature library based on this.
[0054] In step S300, a risk identification model is constructed hierarchically based on the feature library.
[0055] For example, in step S300, the early warning device constructs a risk identification model based on the feature library layer according to the determined user layer, and trains and optimizes the risk identification model.
[0056] The early warning device divides the user groups according to the similarities and differences between different users to determine the user stratification.
[0057] According to some embodiments, the early warning device can divide and stratify the current user groups according to the similarities and differences between different user groups in the upstream and downstream of steel trading. Among them, upstream users can be divided into steel mill users and non-steel mill users according to the supplier's attribution. Downstream users can be divided into steel trader users, sub-end users and end users according to the K-means clustering method.
[0058] Based on the feature library, the early warning device constructs risk identification models according to the determined user layers to output the user's risk occurrence probability.
[0059] According to some embodiments, the early warning device may use a preset algorithm to train the risk identification model through a training data set of a determined risk target variable. The preset algorithm may use a Bagging ensemble learning algorithm.
[0060] Furthermore, the early warning device optimizes the trained risk identification model through a preset parameter optimization method (such as grid search, random search or Bayesian optimization, etc.). The early warning device also calculates the F1 score value (i.e., F1-Scores value) and AUC value through cross-validation to evaluate the performance of the trained risk identification model and verifies it through the training data set.
[0061] The early warning device obtains the probability of risk occurrence through the risk identification model and uses it to set the expected risk loss calculation formula.
[0062] According to some embodiments, the calculation formula of expected risk loss EL can be expressed as follows:
[0063] EL = PD × LGD × EAD, (1)
[0064] Among them, PD is the risk occurrence probability output by the risk identification model, which indicates the probability of a user encountering a risk within a certain period of time. LGD is the loss given default, which indicates the ratio of actual loss to recoverable amount. EAD is the risk exposure amount, which indicates the actual amount corresponding to the user's risk occurrence.
[0065] According to some embodiments, the default loss rate can be obtained through user historical data. Among them, the default loss rate of different categories of users varies with the industry characteristics and economic environment of the users. For example, assuming that the proportion of losses that can be recovered when user risks occur is 30% through user historical data, the default loss rate LGD = 1-30% = 70%.
[0066] In step S400, risk assessment of user real-time data is performed based on the risk identification model.
[0067] For example, in step S400, the early warning device inputs the user's real-time data into the risk identification model to obtain the user's risk occurrence probability and expected risk loss, and generates early warning information.
[0068] The early warning device obtains real-time data from users and inputs it into a risk identification model that has been trained and optimized and verified.
[0069] The early warning device obtains the user's risk occurrence probability through the risk identification model, and obtains the user's expected risk loss through the expected risk loss calculation formula to complete the risk assessment of the user's real-time data.
[0070] According to some embodiments, the early warning device can generate early warning information based on the user's risk probability and expected risk loss according to a preset risk early warning threshold, wherein the early warning information includes the risk level, possible impact range, and recommended control measures.
[0071] Furthermore, the early warning device can push the early warning information to the corresponding business department in the form of a report or dashboard, so that the business department can formulate and implement corresponding risk control measures, track the implementation effect of risk control measures, and make necessary adjustments and optimizations.
[0072] According to the embodiments of the present application, business characteristics and financial analysis can be combined to more accurately identify potential risks and improve the accuracy of risk warnings. A dynamic model update mechanism can be used to ensure that the warning model can adapt to market changes in a timely manner, thereby improving the timeliness of risk warnings. In addition, real-time quantitative assessments can be performed based on market changes, risk estimates of upstream and downstream companies, etc., thereby improving the observability of risk warnings.
[0073] Figure 2 A block diagram of a user risk early warning device according to an exemplary embodiment of the present application is shown.
[0074] like Figure 2 As shown, the early warning device 100 includes a data acquisition module 110 , a feature extraction module 120 , a model construction module 130 and a risk assessment module 140 .
[0075] The data acquisition module 110 acquires user history data from multiple data sources.
[0076] The data acquisition module 110 pre-processes the acquired user history data.
[0077] After preprocessing the user history data, the data acquisition module 110 generates a training data set corresponding to multiple data sources based on the preprocessed user history data.
[0078] The feature extraction module 120 sets a plurality of candidate feature libraries through pre-analysis of steel transaction business processes and financial analysis.
[0079] The feature extraction module 120 determines risk target variables in the training data set.
[0080] According to the determined risk target variable, the feature extraction module 120 determines the data features corresponding to the training data set through a preset feature selection model.
[0081] The feature extraction module 120 uses the screened data features related to the risk target variable as the final data features corresponding to the training data set, and forms a feature library based on them.
[0082] The model building module 130 divides the user groups according to the similarities and differences between different users to determine the user stratification.
[0083] Based on the feature library, the model building module 130 builds risk identification models according to the determined user layers to output the risk occurrence probability of the user.
[0084] The model building module 130 obtains the probability of risk occurrence through the risk identification model and uses it to set the expected risk loss calculation formula.
[0085] The risk assessment module 140 obtains real-time user data and inputs it into a risk identification model that has been trained and optimized and verified.
[0086] The risk assessment module 140 obtains the user's risk occurrence probability through a risk identification model, and obtains the user's expected risk loss through an expected risk loss calculation formula, so as to complete the risk assessment of the user's real-time data.
[0087] The risk assessment module 140 generates warning information according to a preset risk warning threshold based on the user's risk occurrence probability and expected risk loss.
[0088] Figure 3 A block diagram of an electronic device according to an exemplary embodiment of the present application is shown.
[0089] like Figure 3 As shown, the electronic device 600 is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.
[0090] like Figure 3 As shown, the electronic device 600 is in the form of a general-purpose computing device. The components of the electronic device 600 may include, but are not limited to: at least one processing unit 610, at least one storage unit 620, a bus 630 connecting different system components (including the storage unit 620 and the processing unit 610), a display unit 640, etc. The storage unit stores program codes, which can be executed by the processing unit 610, so that the processing unit 610 executes the methods described in this specification according to various exemplary embodiments of the present application. For example, the processing unit 610 may execute the following: Figure 1 The method shown in .
[0091] The storage unit 620 may include a readable medium in the form of a volatile storage unit, such as a random access memory unit (RAM) 6201 and / or a cache memory unit 6202 , and may further include a read-only memory unit (ROM) 6203 .
[0092] The storage unit 620 may also include a program / utility 6204 having a set (at least one) of program modules 6205, such program modules 6205 including but not limited to: an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment.
[0093] Bus 630 may represent one or more of several types of bus structures, including a memory unit bus or memory unit controller, a peripheral bus, an accelerated graphics port, a processing unit, or a local bus using any of a variety of bus architectures.
[0094] The electronic device 600 may also communicate with one or more external devices 700 (e.g., keyboards, pointing devices, Bluetooth devices, etc.), may also communicate with one or more devices that enable a user to interact with the electronic device 600, and / or communicate with any device that enables the electronic device 600 to communicate with one or more other computing devices (e.g., routers, modems, etc.). Such communication may be performed via an input / output (I / O) interface 650. Furthermore, the electronic device 600 may also communicate with one or more networks (e.g., a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) via a network adapter 660. The network adapter 660 may communicate with other modules of the electronic device 600 via a bus 630. It should be understood that, although not shown in the figure, other hardware and / or software modules may be used in conjunction with the electronic device 600, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.
[0095] Through the description of the above implementation methods, it is easy for those skilled in the art to understand that the example embodiments described here can be implemented by software, or by combining software with necessary hardware. The technical solution according to the embodiment of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, including several instructions to enable a computing device (which can be a personal computer, a server, a mobile terminal or a network device, etc.) to execute the method according to the embodiment of the present application.
[0096] The software product may use any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. The readable storage medium may be, for example, but not limited to, a system, device or device of electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination thereof. More specific examples (non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof.
[0097] Computer readable storage media may include data signals propagated in baseband or as part of a carrier wave, wherein readable program codes are carried. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. The readable storage medium may also be any readable medium other than a readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, device, or device. The program codes contained on the readable storage medium may be transmitted using any suitable medium, including but not limited to wireless, wired, optical cable, RF, etc., or any suitable combination thereof.
[0098] Program code for performing the operations of the present application may be written in any combination of one or more programming languages, including object-oriented programming languages such as Java, C++, etc., and conventional procedural programming languages such as "C" or similar programming languages. The program code may be executed entirely on the user computing device, partially on the user device, as a separate software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server. In the case of a remote computing device, the remote computing device may be connected to the user computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computing device (e.g., via the Internet using an Internet service provider).
[0099] The computer-readable medium carries one or more programs. When the one or more programs are executed by a device, the computer-readable medium implements the aforementioned functions.
[0100] Those skilled in the art will appreciate that the above modules can be distributed in the device according to the description of the embodiment, or can be changed accordingly and only used in one or more devices different from the embodiment. The modules of the above embodiments can be combined into one module, or further divided into multiple sub-modules.
[0101] The embodiments of the present application are described in detail above, and the description of the above embodiments is only used to help understand the method and core idea of the present application. At the same time, changes or deformations made by those skilled in the art based on the idea of the present application, the specific implementation method and the scope of application of the present application, all belong to the scope of protection of the present application. In summary, the content of this specification should not be construed as limiting the present application.
Claims
1. A user risk early warning method, characterized in that: include: Generate training data sets based on user historical data from multiple data sources; Extracting data features corresponding to the training data set to form a feature library; Based on the feature library, a risk identification model is constructed hierarchically; Based on the risk identification model, risk assessment of user real-time data is performed.
2. The method according to claim 1, characterized in that Generate training datasets based on historical user data from multiple data sources, including: Obtaining the user's historical data; Preprocessing the user historical data; Based on the pre-processed user history data, a training data set corresponding to the multiple data sources is generated.
3. The method according to claim 1, characterized in that Extracting data features corresponding to the training data set to form a feature library includes: Setting multiple candidate feature libraries; determining a risk target variable in the training data set; According to the risk target variable, determining the data features corresponding to the training data set through a preset feature selection model; The feature library is determined from among the plurality of candidate feature libraries based on data features corresponding to the training data set.
4. The method according to claim 1, characterized in that: Based on the feature library, a risk identification model is constructed hierarchically, including: Divide user groups to determine user stratification; According to the user stratification, the risk identification model is constructed based on the feature library stratification to output the risk occurrence probability of the user; Training the risk identification model using the training data set; and The risk identification model that has been trained is optimized and verified.
5. The method according to claim 4, characterized in that The method further comprises: According to the probability of occurrence of the risk, a formula for calculating the expected risk loss is set.
6. The method according to claim 5, characterized in that According to the probability of occurrence of the risk, a formula for calculating the expected risk loss is set, including: Obtaining default loss rate through the user historical data; Get the actual amount corresponding to the user's risk when it occurs; The expected risk loss calculation formula is set according to the risk occurrence probability, the default loss rate and the actual amount corresponding to the user risk when it occurs.
7. The method according to claim 1, characterized in that Based on the risk identification model, risk assessment of user real-time data is performed, including: Inputting the user real-time data into the risk identification model; Obtain the user's risk occurrence probability and expected risk loss through the risk identification model and expected risk loss calculation formula; Based on the risk occurrence probability and the expected risk loss, warning information is generated according to a preset risk warning threshold.
8. A user risk early warning device, characterized in that: include: A data acquisition module, used to generate a training data set based on user historical data from multiple data sources; A feature extraction module, used to extract data features corresponding to the training data set to form a feature library; A model building module, used for hierarchically building a risk identification model based on the feature library; The risk assessment module is used to perform risk assessment on user real-time data based on the risk identification model.
9. An electronic device, characterized in that: include: one or more processors; A storage device for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 7.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.