A credit investigation report analysis method and system, a terminal device and a storage medium

Abstract

The application provides a credit investigation report analysis method and system, terminal equipment and a storage medium, constructs credit investigation report features and a credit investigation report feature group, and reports credit investigation report identifiers in a database, takes out corresponding credit investigation reports, calculates the credit investigation report features according to the taken out credit investigation reports, filters, marks and calculates all credit investigation reports in a folder where the credit investigation reports are located, creates a model data set, and saves the model data set to the database. The same feature can obtain user information from credit investigation reports in different formats, and supports filtering, conversion and function calculation of user information according to specific requirements. The feature can be developed and calculated online according to business, the credit investigation report analysis is more flexible, automatic and intelligent, and the credit investigation report analysis efficiency is improved. The problem that different styles of credit investigation report files are difficult to analyze in the prior art is solved.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of financial services, and relates to a credit report analysis method and system, a terminal device and a storage medium. BACKGROUND

[0002] In financial loan business, credit reports play a decisive role. All loan businesses use credit reports for risk assessment, and develop their own businesses according to the assessment reports. Therefore, each financial institution will establish its own credit system to obtain, analyze and store credit data, which causes the diversity of credit reports. For example, in personal credit reports, there are different document structures such as JavaScript object Notation (JSON), Extensible Markup Language (XML) and HTML. This causes great difficulty to the important module of report analysis. At present, most report analysis work is only for fixed format credit reports. How to realize the analysis of credit report files of various styles and obtain user information is a problem to be solved. SUMMARY

[0003] The application aims to solve the problems in the prior art and provide a credit report analysis method, system, terminal device and storage medium.

[0004] To achieve the above-mentioned purpose, the application adopts the following technical solutions:

[0005] A credit report analysis method comprises the following steps:

[0006] S1: receiving information related to the construction of credit report features, constructing credit report features, and storing the constructed credit report features in a database;

[0007] S2: grouping the credit report features, creating a credit report feature group, and storing the created credit report feature group in the database;

[0008] S3: taking out the corresponding credit report according to the credit report identifier in the credit report database, and calculating the credit report features according to the taken out credit report;

[0009]

[0009] S4: receiving the folder where the credit report is located and the credit report feature group, filtering and marking all credit reports in the folder, calculating all credit report features, creating a model data set, and saving the model data set to the database.

[0010] The application is further improved in that:

[0011] The S1 comprises the following steps:

[0012] Receiving the input credit report feature English name and Chinese name, wherein the Chinese name is used as basic information identification, and the English name is used as credit report feature unique identification;

[0013] The received information is packaged into an object, and the object is serialized into a JSON string and persisted to a database.

[0014] The S3 includes the following steps:

[0015] A class file for credit report deserialization is generated, the credit report is deserialized, converted into a CreditReport object, and converted into an ObjectTree structure object, data is extracted from the ObjectTree, and the extracted data is mapped and filtered to calculate the final result.

[0016] In the S3, the method for generating a class file for credit report deserialization is:

[0017] The folder or file name of the credit report is passed in and deserialized, the deserialized object is converted into a StructNode structure, all credit report trees are merged into a monocular tree, the merged monocular tree is alias mapped according to the hierarchical structure, a configuration file is read, and the Package corresponding to the generated class file is obtained.

[0018] The S4 includes the following steps:

[0019] The document of the credit report is received, all credit report features received are divided into two parts, namely credit report business features and credit report marker features, all credit report features are calculated in turn, after the calculation is completed, the features are re-split into business feature groups and marker features, then the calculation results are packaged in JSON format, the calculated credit report is marked, the final packaged results are serialized into a JSON string, and uploaded to a database.

[0020] The calculation result of the credit report business feature corresponds to the feature value of the credit report in the business, serving as the main body of the data set; the calculation result of the credit report marker feature serves as the marking result of the credit report.

[0021] A credit report analysis system includes a credit report feature module, a credit report feature group module, a credit report feature calculation module, and a model data set module.

[0022] The credit report feature module is used to receive relevant information for building credit report features, build credit report features, and store the built credit report features in a database.

[0023] The credit report feature group module is configured to group credit report features, create credit report feature groups, and store the created credit report feature groups in a database.

[0024] The credit report feature calculation module is configured to take out corresponding credit reports according to credit report identifiers in the credit report database, and calculate credit report features according to the taken out credit reports.

[0025] The model data set module is configured to receive a folder in which credit reports are located and credit report feature groups, filter and mark all credit reports in the folder, calculate all credit report features, create a model data set, and save the model data set to a database.

[0026] A terminal device comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the steps of the method according to any one of the present application when executing the computer program.

[0027] A computer readable storage medium stores a computer program, and the computer program implements the steps of the method according to any one of the present application when executed by a processor.

[0028] Compared with the prior art, the present application has the following beneficial effects:

[0029] The present application proposes a credit report analysis method, which analyzes and calculates credit report features by constructing credit report features and credit report feature groups, realizes that the same feature can obtain user information from credit reports of different formats, supports filtering, conversion and function calculation of user information according to specific requirements, and enables features to be developed and calculated online according to business, so that the analysis of credit reports is more flexible, automated and intelligent, and the analysis efficiency of credit reports is improved.

[0030] Further, the class file for analyzing credit report files can be dynamically generated in the present application, so that different formats of credit reports can be dynamically analyzed. Then, through alias mapping, the problem of inconsistent label names of credit reports of different institutions is solved. BRIEF DESCRIPTION OF DRAWINGS

[0031] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments. It should be understood that the following drawings only show some embodiments of the present application, and therefore should not be regarded as a limitation on the scope, and for those skilled in the art, other related drawings can also be obtained without creative labor on the basis of these drawings;

[0032] Figure 1This is a flowchart illustrating the credit report feature construction process of an embodiment of the present invention.

[0033] Figure 2 This is a graph illustrating the calculation of credit report features according to an embodiment of the present invention.

[0034] Figure 3 This is a flowchart illustrating the class file construction process of an embodiment of the present invention.

[0035] Figure 4 The dataset creation process for this embodiment of the invention. Detailed Implementation

[0036] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0037] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0038] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0039] In the description of the embodiments of the present invention, it should be noted that if terms such as "upper," "lower," "horizontal," or "inner" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, or the orientation or positional relationship commonly used when the product of the invention is in use, they are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the present invention. Furthermore, terms such as "first" and "second" are only used to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0040] Furthermore, the use of the term "horizontal" does not imply that the component must be absolutely horizontal, but rather that it can be slightly tilted. For example, "horizontal" simply means that its direction is more horizontal than "vertical," and does not mean that the structure must be completely horizontal, but can be slightly tilted.

[0041] In the description of the embodiments of the present application, it also needs to be explained that, unless otherwise explicitly specified and limited, if the terms "arrange", "install", "connect", "join" appear, they should be understood in a broad sense, for example, can be fixedly connected, can also be detachably connected, or integrally connected; can be mechanically connected, can also be electrically connected; can be directly connected, can also be indirectly connected through an intermediate medium, can be the communication inside two elements. For those skilled in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.

[0042] The present application will be further described in detail below with reference to the accompanying drawings:

[0043] Referring to Figure 1 The embodiment of the present application discloses a credit report analysis method, comprising the following steps:

[0044] S1: receiving the related information of constructing the credit report features, constructing the credit report features, storing the constructed credit report features into the database, providing query and modification functions;

[0045] Among them, the credit report features are divided into three parts, which are data acquisition, data processing and data calculation. After receiving the related configuration information, it is encapsulated as an object and serialized as a JSON string, and persisted to the database;

[0046] Referring to Figure 1 , specifically comprising the following steps:

[0047] Step S101, receiving the input credit report feature English name and Chinese name as basic information identification, the feature English name is used as the unique identification of the credit report feature in the present application.

[0048] Step S102, encapsulating the path where the required data is located. The path is used as the path identification in the present application. The value of the path is a string obtained by splicing the node names of all nodes passed from the root node to the target node, and / is used as the separator between all node names.

[0049] Step S103, encapsulating the pre-processing operation before filtering. Create the function needed for pre-processing, and add pre-processing operation to the column that needs to be pre-processed. The pre-processing function of the present application is realized by function call.

[0050] Step S104, encapsulating the filtering operation. The present application refers to the filtering operation in the database, which can use greater than, greater than or equal to, less than, less than or equal to, equal to, in, not in and other keywords, and filter the entire two-dimensional table according to the data of a column;

[0051] Step S105, encapsulate the calculation operation. Create a function needed for the calculation operation to calculate the final result of the feature. The calculation function of the present application is implemented by function service call method. The input of the function is multiple lists, and the output of the function is a string as the final calculation result.

[0052] Step S106, perform serialization operation on all encapsulated objects to obtain a JSON string for return, that is, the created feature.

[0053] S2: According to the business logic, the credit investigation report features are grouped, the credit investigation report feature groups are created, the created credit investigation report feature groups are stored in the database, and the query and modification functions are provided;

[0054] S3: According to the credit investigation report identifier in the credit investigation report database, the corresponding credit investigation report is taken out, and the credit investigation report features are calculated according to the credit investigation report and the credit investigation report features;

[0055] It includes receiving the credit investigation report identifier or credit investigation report file and the credit investigation report feature identifier or credit investigation report feature JSON string, completing the calculation of the credit investigation report features, and returning the calculation result.

[0056] Receive the credit investigation report identifier or credit investigation report and the credit investigation report feature group identifier or credit investigation report feature group JSON string, complete the calculation of a feature group, and return the result in the format of JSON string.

[0057] The credit investigation report is deserialized to generate a Java object.

[0058] The Java object is converted into a fixed structure ObjectTree structure. The ObjectTree structure has four attributes, namely node name, ObjectChildren, ListChildren and data.

[0059] At the same time of ObjectTree structure conversion, the data is mapped to map some numerical data to corresponding Chinese data of the business.

[0060] Referring to Figure 2 , specifically comprising the following steps:

[0061] Step S301, according to the credit investigation report file, a class file for deserializing the credit investigation report is generated. Different types of credit investigation reports will adopt different strategies. The types of credit investigation reports supported in the present application are three, which are XML, JSON and Excel hierarchical structure model.

[0062] Step S302, the credit report is deserialized and converted into a CreditReport object. Different types of credit reports will adopt different strategies. There are two formats of credit reports supported in the present application, which are XML and JSON.

[0063] Step S303, the CreditReport object is converted into an ObjectTree structure object. The ObjectTree structure has four attributes, which are node name, ObjectChildren, ListChildren and data. The present application uses a recursive algorithm to convert the CreditReport object into an ObjectTree object. In the conversion process, all child nodes are divided into two categories, which are objects and arrays. A HashMap is used as a data structure, the name of the object as the key of the map, and the converted Tree as the value of the map. At the same time of conversion, data mapping is performed to map some numerical identification data to the corresponding Chinese meaning, which is convenient for understanding and operation.

[0064] Step S304, data retrieval operation. The present application uses a recursive algorithm to retrieve data from the ObjectTree according to the path encapsulated in the data retrieval logic. The data retrieval process of the present application is divided into three cases: first, path interruption occurs. When searching for data through the path, if the corresponding node cannot be found during the path search process, the path interruption phenomenon occurs, which represents that the path data does not exist, and a specific non-existent identifier is returned directly. Second, the path does not interrupt, and the last node is a leaf node, then the data of the leaf node is returned directly as the search result. Third, the path does not interrupt, but the last node is not a leaf node, then the last node is serialized as a whole and converted into a JSON string, and the JSON string is taken as the final result. Finally, the retrieved data is encapsulated into a two-dimensional table, taking the identifier of each path as the column name.

[0065] Step S305, preprocessing operation. According to the filtering preprocessing logic encapsulated in the credit report features, the column data is preprocessed. The present application uses a function call scheme to realize the preprocessing operation. The preprocessing operation contains function name and column name, and the function calculation service is called to calculate the function name and parameters for preprocessing operation. The function calculation service in the present application provides Java and python function calculation functions. First, the column data is taken from the two-dimensional table according to the column name, and the data is converted, using comma as the separator, to convert the List into a string. Then the preprocessing operation is performed on the function calculation service side, and a string separated by commas is returned, which is split into a List as new data. The old data in the table is deleted and the new data is added.

[0066] Step S306, filtering operation. The two-dimensional table is filtered according to the filtering logic encapsulated in the credit report features. The present application refers to the filtering operation in the Mysql database, and realizes filtering with greater than, greater than or equal to, less than, less than or equal to, equal to, in, not in and other logic. First, according to the column name, the column to be filtered is taken, the column data is traversed, and it is judged according to the encapsulated filtering logic whether the data meets the condition, and if it does not meet the condition, the row data is deleted in the two-dimensional table.

[0067] Step S307, calculation operation. According to the calculation logic encapsulated in the credit report features, the final result is calculated. The present application adopts the form of calling functions to complete the calculation operation. The features encapsulate function names and column names, call the function calculation service, and pass in the function name and parameter to perform the calculation operation. The function calculation service in the present application provides the functions of Java and python function calculation. First, all column data in the two-dimensional table is taken, and data transformation is performed on each column data. A string is formed by taking a comma as a separator. Then all the column data is transformed into a JSON string as a parameter by taking the column name as the key. On the function calculation server, the function entry is located according to the function name, and the calculation is performed on the input parameter to return a string as the calculation result.

[0068] Step S308, return the calculation result. The result is returned in the format of a JSON string. The credit report feature name is taken as the key, and the calculation result is taken as the value.

[0069] Reference Figure 3 In the present embodiment, the creation process of the S301 class file is as follows:

[0070] Step S301-1, the folder where the credit report is located or the file name is passed in. If the input file is a folder, the file deserialization tool is used to deserialize all credit reports in the folder. If it is a file, the file is deserialized. The present application supports two file types, which are JSON and XML.

[0071] Step S301-2, the object generated by the first step of deserialization is converted into a StructNode structure. The StructNode in the present application contains four attributes, namely the node name, the node alias, the node type, and the child node. The data type of the child node is List, and the generic type of List is StructNode. The conversion logic of the present application is as follows: if the child of the current object is an object type, the element object converted StructNode is directly retained as the child node, and the node type is set to Object; if the child of the current object is an array type, the StructNode converted from each element in the array is taken as the child node, and the node type is modified to List.

[0072] Step S301-3, tree merging. The Trees obtained from all credit reports are merged into a monadic tree. In the present application, the meaning of a monadic tree is that any node does not have two child nodes with the same name. First, each StructNode in the List is merged by itself, and all StructNodes are converted into a monadic tree. The processing logic of the present application is as follows: if one child node of the node has a name different from all other child nodes, the child node is directly retained, and the type of the node is modified to Object. If two or more child nodes in the node have the same name, all child nodes with the same name are merged to obtain a node, and the type of the node is modified to List, and a monadic tree can be obtained after merging. Then, all merged monadic trees are traversed from the second tree, and merged with the first tree. In the present application, the merging logic is as follows: the roles of the two trees are positioned as a root tree and a resource tree, and the nodes in the resource tree are used to supplement the nodes of the root tree. If the resource tree contains a node that the tree does not have, the node is directly added to the root tree. If there is a node with the same name, the nodes are recursively merged. At the same time, the node type of the root tree is modified according to the node type of the resource tree, and the modification rule is List>Object>String.

[0073] Step S301-4, alias mapping. The merged monadic tree is mapped according to the hierarchical structure. In the present application, the alias of a node is mapped according to the hierarchical access path of the node in the tree. When writing a class file, the alias is used as the attribute of the class file, and the node name is used as the name of the deserialization file, forming a one-to-one mapping relationship

[0074] Step S301-5, writing a class file. The configuration file is read to obtain the Package corresponding to the generated class file. The present application uses a depth-first traversal algorithm and uses a Velocity template technology to generate a corresponding class file for each node. The class name is the hierarchical access path of the node, and the attributes of the class are all child nodes of the node. The class file generated in the present application needs to be used in conjunction with the lombok plug-in.

[0075] S4: receiving the folder of credit report and the feature group of credit report, filtering and marking all credit reports in the folder, creating a model data set, and saving the model data set to a database.

[0076] The incoming credit report features are divided into two parts, namely the business features of the credit report and the marked features of the credit report. After calculating all the credit report features, the calculation result of the business features corresponds to the feature value of the credit report in the business, which is the main part of the data set. The calculation result of the marked features is processed according to the corresponding strategy to obtain the marking result of the credit report.

[0077] Referring to Figure 4 , step S401, the document of the incoming credit report. All credit reports in the document library are marked.

[0078] Step S402, the business feature group of the incoming credit report and the marked feature of the credit report.

[0079] Step S403, calculate all features. Put the business feature group and the marked feature into the Set collection to prevent repeated calculation. Then call the feature calculation service, input the credit report and the feature collection, and complete the calculation of the entire feature collection.

[0080] Step S404, encapsulate a single data set in JSON format. After completing the calculation of all feature collections, further split the features into a business feature group and a marked feature. Then encapsulate the calculation result in JSON format. The format is {“data”:{},”result”:{}}. The data encapsulates all the values of the business feature group. The result is the calculation value of the marked feature.

[0081] Step S405, loop marking all credit reports in the document and encapsulate them. Query all credit reports in the document, loop 503 and 504 steps, complete the calculation of all credit reports. Finally, encapsulate the result in an array format into a JSON string.

[0082] Step S406, serialize the final encapsulated result into a JSON string, upload it to the database, and provide a download function.

[0083] The embodiment of the application also discloses a credit report analysis system, comprising a credit report feature module, a credit report feature group module, a credit report feature calculation module and a model data set module

[0084] The credit investigation report feature module is configured to receive relevant information for constructing credit investigation report features, construct credit investigation report features, and store the constructed credit investigation report features in a database.

[0085] The credit investigation report feature group module is configured to group credit investigation report features, create credit investigation report feature groups, and store the created credit investigation report feature groups in a database.

[0086] The credit investigation report feature calculation module is configured to take out corresponding credit investigation reports according to credit investigation report identifiers in the credit investigation report database, and calculate credit investigation report features according to the taken out credit investigation reports.

[0087] The model data set module is configured to receive credit investigation report folders and credit investigation report feature groups, filter and mark all credit investigation reports in the folders, calculate all credit investigation report features, create a model data set, and save the model data set to a database.

[0088] According to the embodiment of the present application, the class construction module can dynamically generate a class file for analyzing credit investigation report files, so as to dynamically analyze credit investigation reports of different formats. Then, the alias mapping solves the problem of inconsistent label names of credit investigation reports of different institutions. The feature construction is performed from three parts of data acquisition, data processing and data calculation, so that the same feature can acquire user information from credit investigation reports of different formats, and support filtering, conversion and function calculation of user information according to specific requirements. Meanwhile, the feature can be developed and calculated online according to the business, so that the analysis of the credit investigation report is more flexible, automatic and intelligent, and the analysis efficiency of the credit investigation report is improved.

[0089] An embodiment of the present application provides a schematic diagram of a terminal device. The terminal device of the embodiment comprises a processor, a memory, and a computer program stored in the memory and executable on the processor. The processor implements the steps in each of the method embodiments when executing the computer program. Alternatively, the processor implements the functions of each module / unit in each of the device embodiments when executing the computer program.

[0090] The computer program can be divided into one or more modules / units, which are stored in the memory and executed by the processor to complete the present application.

[0091] The terminal device can be a desktop computer, a notebook computer, a palm computer, a cloud server and other computing devices. The terminal device can include, but is not limited to, a processor and a memory.

[0092] The processor can be a central processing unit (CPU), and can also be other general-purpose processors, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic device, discrete hardware component, etc.

[0093] The memory can be used to store the computer program and / or modules, and the processor realizes various functions of the terminal device by running or executing the computer program and / or modules stored in the memory, and calling the data stored in the memory.

[0094] The modules / units integrated in the terminal device, if realized in the form of software function units and sold or used as independent products, can be stored in a computer readable storage medium. Based on such understanding, all or part of the processes in the above-mentioned embodiment methods can also be completed by a computer program instructing related hardware, and the computer program can be stored in a computer readable storage medium. The computer program can realize the steps of the above-mentioned various method embodiments when executed by a processor. The computer program includes computer program code, which can be in the form of source code, object code, executable file or some intermediate form. The computer readable medium can include any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal and software distribution medium, etc. It should be noted that the computer readable medium can include or exclude contents according to the requirements of legislation and patent practice in the jurisdiction, for example, in some jurisdictions, according to legislation and patent practice, the computer readable medium does not include electrical carrier signals and telecommunication signals.

[0095] The above only describes the preferred embodiments of the present application and is not used to limit the present application. For those skilled in the art, the present application can have various modifications and changes. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for parsing credit reports, characterized in that, Includes the following steps: S1: Receive relevant information for constructing credit report features, construct credit report features, and store the constructed credit report features in the database; S2: Group the credit report features, create credit report feature groups, and store the created credit report feature groups in the database; S3: Based on the credit report identifier in the credit report database, retrieve the corresponding credit report, and calculate the credit report characteristics based on the retrieved credit report; S4: Receive the folder containing the credit reports and the feature groups of the credit reports, filter and mark all credit reports in the folder, calculate the features of all credit reports, create a model dataset, and save the model dataset to the database; S3 includes the following steps: Generate a class file for deserializing the credit report, deserialize the credit report, convert it into a CreditReport object, convert the CreditReport object into an ObjectTree structure object, extract data from the ObjectTree, and simultaneously map and filter the extracted data and calculate the final result. In step S3, the method for generating the deserialized class file of the credit report is as follows: The program takes the folder or file name of the credit report as input, deserializes it, converts the deserialized object into a StructNode structure, merges the Trees obtained from all the credit reports into a monocular tree, performs alias mapping on the merged monocular tree according to the hierarchical structure, reads the configuration file, and obtains the Package corresponding to the generated class file. S4 includes the following steps: The document containing the credit report is received. All credit report features are divided into two parts: credit report business features and credit report tag features. All credit report features are calculated sequentially. After the calculation is completed, the features are split back into business feature groups and tag features. The calculation results are then encapsulated in JSON format. At the same time, the calculated credit report is tagged. The final encapsulated result is serialized into a JSON string and uploaded to the database. The calculation results of the credit report business features correspond to the feature values ​​of the credit report in the business and serve as the main body of the dataset; the calculation results of the credit report labeling features serve as the labeling results of the credit report.

2. The credit report parsing method according to claim 1, characterized in that, S1 includes the following steps: The system receives the English and Chinese names of the credit report features, where the Chinese name serves as the basic information identifier and the English name serves as the unique identifier for the credit report feature. The received information is encapsulated into an object, and the object is serialized into a JSON string and persisted to the database.

3. A credit report parsing system based on the method of claim 1, characterized in that, It includes a credit report feature module, a credit report feature group module, a credit report feature calculation module, and a model dataset module; The credit report feature module is used to receive relevant information for constructing credit report features, construct credit report features, and store the constructed credit report features in the database. The credit report feature group module is used to group credit report features, create credit report feature groups, and store the created credit report feature groups in the database; The credit report feature calculation module is used to retrieve the corresponding credit report based on the credit report identifier in the credit report database, and calculate the credit report features based on the retrieved credit report. The model dataset module is used to receive the folder containing the credit reports and the feature groups of the credit reports, filter and label all credit reports in the folder, calculate the features of all credit reports, create a model dataset, and save the model dataset to the database.

4. A terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1-2.

5. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1-2.

Images