Risk factor-based grade calculation method, device, equipment and storage medium
By structuring the historical business data and analyzing the grayscale correlation algorithm, the child and parent factors of disaster risk are extracted and calculated, and the problem of low accuracy of risk level calculation in the existing technology is solved, and the matching of disaster risk level is improved accuracy.
Patent Information
- Application Number
- CN202111196116.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-10-14
- Publication Date
- 2024-12-03
- Estimated Expiration
- 2041-10-14
AI Technical Summary
In the prior art, disaster risk level calculation relies on single or independent risk factors and cannot conduct comprehensive correlation analysis, resulting in low accuracy of risk level calculation.
By obtaining the historical basic business data and historical execution business data of the comprehensive business area, performing structured processing, extracting the target risk sub-factors and parent factors based on the grayscale association algorithm, generating the target correlation data set, and calculating the target risk with the risk query instructions. grade.
The accuracy of calculation correlation data is improved, thereby improving the accuracy of matching target risk levels and saving server computing resources.
Smart Images

Figure CN113869780B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of data processing, and in particular to a risk factor-based level calculation method, device, equipment and storage medium. Background Art
[0002] Many areas will suffer from various disasters, such as floods, fires, etc. Such disasters will have a serious impact on people's property. Therefore, many companies will launch some guaranteed products for disasters to compensate for damaged items. How to accurately measure the risk level of items in disasters is an issue that companies urgently need to solve.
[0003] In the prior art, the disaster risk level of a specified area is mainly calculated using a risk calculation model of a single or independent risk factor. Since the choice of risk factors is relatively single, a comprehensive correlation analysis cannot be performed, resulting in a low accuracy rate in generating risk levels. Summary of the invention
[0004] The present invention provides a risk factor-based level calculation method, device, equipment and storage medium, which are used to improve the accuracy of calculating correlation data, thereby improving the accuracy of matching target risk levels.
[0005] The first aspect of the present invention provides a risk factor-based grade calculation method, comprising: obtaining historical basic business data and historical execution business data of the comprehensive business area, and structuring the historical basic business data and the historical execution business data to generate structured historical business data; generating a target risk factor based on the structured historical business data and a preset risk factor model; extracting multiple target risk sub-factors and multiple target risk parent factors from the structured historical business data based on a grayscale association algorithm and the target risk factor, and performing calculations based on the multiple target risk sub-factors and the multiple target risk parent factors to generate a target correlation data set, wherein the target risk sub-factors and the target risk parent factors correspond one to one; obtaining a risk query instruction, performing risk grade calculation based on the risk query business identifier in the risk query instruction in combination with the target correlation data set and the structured historical business data to generate a target risk grade.
[0006] Optionally, in a first implementation method of the first aspect of the present invention, the historical basic business data of the integrated business area and the historical execution business data of the integrated business area are obtained, and the historical basic business data and the historical execution business data are structured to generate structured historical business data, including: obtaining the historical basic business data of the integrated business area and the historical execution business data of the integrated business area; reading the sub-business area code corresponding to each sub-business area from the historical basic business data and the historical execution business data to obtain multiple sub-business area codes, and reading the business object code corresponding to each business object from the historical basic business data and the historical execution business data to obtain multiple business object codes, wherein the integrated business area includes multiple sub-business areas and multiple business objects; and structured processing the historical basic business data and the historical execution business data based on the multiple sub-business area codes and the multiple business object codes to generate structured historical business data.
[0007] Optionally, in a second implementation method of the first aspect of the present invention, the structuring processing of the historical basic business data and the historical execution business data based on the multiple sub-business area codes and the multiple business object codes to generate structured historical business data includes: obtaining cleanup instructions, and performing data cleansing on the historical basic business data and the historical execution business data in accordance with the cleanup instructions in combination with the multiple sub-business area codes and the multiple business object codes to generate cleaned historical basic business data and cleaned historical execution business data; structuring the cleaned historical basic business data and the cleaned historical execution business data in combination with the multiple sub-business area codes and the multiple business object codes to generate structured historical business data.
[0008] Optionally, in a third implementation method of the first aspect of the present invention, generating a target risk factor based on the structured historical business data and a preset risk factor model includes: calling a preset random function to randomly select multiple initial risk factors from the structured historical business data; inputting the multiple initial risk factors into the preset risk factor model for weight calculation, generating business weight parameters corresponding to each initial risk factor, and obtaining multiple business weight parameters; determining the largest business weight parameter from the multiple business weight parameters as the target business weight parameter, and determining the initial risk factor corresponding to the target business weight parameter as the target risk factor.
[0009] Optionally, in a fourth implementation manner of the first aspect of the present invention, the grayscale association algorithm and the target risk factor are used to extract multiple target risk sub-factors and multiple target risk parent factors from the structured historical business data, and calculations are performed based on the multiple target risk sub-factors and the multiple target risk parent factors to generate a target correlation data set, and the one-to-one correspondence between the target risk sub-factors and the target risk parent factors includes: calling a preset sub-factor extraction function to extract sub-factors from the structured historical business data based on the target risk factor to generate multiple target risk sub-factors; calling a preset parent extraction function to extract parent factors from the structured historical business data based on the target risk factor and each target risk sub-factor to generate a target risk parent factor corresponding to each target risk sub-factor, and obtain multiple target risk parent factors; performing correlation calculations on each target risk sub-factor and the corresponding target risk parent factor based on the grayscale association algorithm to generate initial correlation data corresponding to each target risk sub-factor, and obtain an initial correlation data set; sorting multiple initial correlation data in the initial correlation data set in order from large to small to generate a target correlation data set, and the target correlation data set includes multiple target correlation data.
[0010] Optionally, in a fifth implementation manner of the first aspect of the present invention, the risk query instruction is obtained, and the risk level is calculated according to the risk query business identifier in the risk query instruction in combination with the target correlation data set and the structured historical business data to generate a target risk level, including: obtaining a risk query instruction, parsing the risk query instruction, generating a risk query business identifier, and performing a factor query in the structured historical business data based on the risk query business identifier to generate risk sub-factors to be queried; classifying the target correlation data set according to preset grade classification rules to generate multiple risk levels; reading the target correlation data corresponding to the risk sub-factor to be queried in the target correlation data set, and matching the risk levels based on the target correlation data to obtain the target risk level.
[0011] Optionally, in a sixth implementation manner of the first aspect of the present invention, when obtaining a risk query instruction, calculating the risk level according to the risk query business identifier in the risk query instruction in combination with the target correlation data set and the structured historical business data, and generating the target risk level, the risk factor-based level calculation method also includes: transmitting the target risk level corresponding to the risk query instruction to a terminal corresponding to the risk query instruction.
[0012] The second aspect of the present invention provides a risk factor-based level calculation device, including: an acquisition module, used to acquire historical basic business data of a comprehensive business area and historical execution business data of the comprehensive business area, and perform structured processing on the historical basic business data and the historical execution business data to generate structured historical business data; a risk factor generation module, used to generate a target risk factor based on the structured historical business data and a preset risk factor model; a calculation module, used to extract multiple target risk sub-factors and multiple target risk parent factors from the structured historical business data based on a grayscale association algorithm and the target risk factor, and perform calculations based on the multiple target risk sub-factors and the multiple target risk parent factors to generate a target correlation data set, with the target risk sub-factors and the target risk parent factors corresponding one to one; a query module, used to acquire a risk query instruction, perform risk level calculation based on the risk query business identifier in the risk query instruction in combination with the target correlation data set and the structured historical business data, and generate a target risk level.
[0013] Optionally, in a first implementation method of the second aspect of the present invention, the acquisition module includes: an acquisition unit, used to acquire historical basic business data of the integrated business area and historical execution business data of the integrated business area; a reading unit, used to read the sub-business area code corresponding to each sub-business area from the historical basic business data and the historical execution business data to obtain multiple sub-business area codes, and read the business object code corresponding to each business object from the historical basic business data and the historical execution business data to obtain multiple business object codes, wherein the integrated business area includes multiple sub-business areas and multiple business objects; a structuring unit, used to structure the historical basic business data and the historical execution business data based on the multiple sub-business area codes and the multiple business object codes to generate structured historical business data.
[0014] Optionally, in a second implementation method of the second aspect of the present invention, the structuring unit may also be specifically used to: obtain a cleanup instruction, and perform data cleanup on the historical basic business data and the historical execution business data in accordance with the cleanup instruction in combination with the multiple sub-business area codes and the multiple business object codes to generate the historical basic business data after data cleansing and the historical execution business data after data cleansing; perform structured processing on the historical basic business data after data cleansing and the historical execution business data after data cleansing in combination with the multiple sub-business area codes and the multiple business object codes to generate structured historical business data.
[0015] Optionally, in a third implementation method of the second aspect of the present invention, the risk factor generation module can also be specifically used to: call a preset random function to randomly select multiple initial risk factors from the structured historical business data; input the multiple initial risk factors into a preset risk factor model for weight calculation, generate business weight parameters corresponding to each initial risk factor, and obtain multiple business weight parameters; determine the largest business weight parameter from the multiple business weight parameters as the target business weight parameter, and determine the initial risk factor corresponding to the target business weight parameter as the target risk factor.
[0016] Optionally, in a fourth implementation manner of the second aspect of the present invention, the calculation module can also be specifically used to: call a preset sub-factor extraction function to extract sub-factors from the structured historical business data based on the target risk factor, and generate multiple target risk sub-factors; call a preset parent function to extract parent factors from the structured historical business data based on the target risk factor and each target risk sub-factor, and generate a target risk parent factor corresponding to each target risk sub-factor, and obtain multiple target risk parent factors; calculate the correlation between each target risk sub-factor and the corresponding target risk parent factor based on the grayscale correlation algorithm, generate initial correlation data corresponding to each target risk sub-factor, and obtain an initial correlation data set; sort multiple initial correlation data in the initial correlation data set in order from large to small, and generate a target correlation data set, which includes multiple target correlation data.
[0017] Optionally, in the fifth implementation method of the second aspect of the present invention, the query module can also be specifically used to: obtain a risk query instruction, parse the risk query instruction, generate a risk query business identifier, and perform a factor query in the structured historical business data based on the risk query business identifier to generate risk sub-factors to be queried; grade the target correlation data set according to preset grade classification rules to generate multiple risk levels; read the target correlation data corresponding to the risk sub-factor to be queried in the target correlation data set, and match the risk levels based on the target correlation data to obtain the target risk level.
[0018] Optionally, in a sixth implementation of the second aspect of the present invention, the risk factor-based level calculation device further includes: a transmission module, configured to transmit the target risk level corresponding to the risk query instruction to a terminal corresponding to the risk query instruction.
[0019] The third aspect of the present invention provides a risk factor-based grade calculation device, comprising: a memory and at least one processor, wherein the memory stores instructions; the at least one processor calls the instructions in the memory so that the risk factor-based grade calculation device executes the above-mentioned risk factor-based grade calculation method.
[0020] A fourth aspect of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores instructions, which, when executed on a computer, enable the computer to execute the above-mentioned risk factor-based grade calculation method.
[0021] In the technical solution provided by the present invention, historical basic business data of the comprehensive business area and historical execution business data of the comprehensive business area are obtained, and the historical basic business data and the historical execution business data are structured to generate structured historical business data; a target risk factor is generated according to the structured historical business data and a preset risk factor model; based on the grayscale association algorithm and the target risk factor, multiple target risk sub-factors and multiple target risk parent factors are extracted from the structured historical business data, and calculations are performed based on the multiple target risk sub-factors and the multiple target risk parent factors to generate a target correlation data set, in which the target risk sub-factors and the target risk parent factors correspond one to one; a risk query instruction is obtained, and the risk level calculation is performed according to the risk query business identifier in the risk query instruction in combination with the target correlation data set and the structured historical business data to generate a target risk level. In an embodiment of the present invention, historical basic business data and historical execution business data are structured into structured historical business data, and then a target risk factor and a correlation data set are generated based on a risk factor model and a grayscale correlation algorithm. Finally, a target risk level corresponding to the risk query instruction is generated according to the risk query instruction and the correlation data set. The structured historical business data saves computing resources of the server. In business usage scenarios, the target risk factor is determined in combination with the risk factor model, and then a correlation data set is generated based on the target risk factor, thereby improving the accuracy of calculating the correlation data and thus improving the accuracy of matching the target risk level. BRIEF DESCRIPTION OF THE DRAWINGS
[0022] Figure 1 A schematic diagram of an embodiment of a method for calculating a level based on risk factors in an embodiment of the present invention;
[0023] Figure 2 It is a schematic diagram of another embodiment of the method for calculating the level based on the risk factor in an embodiment of the present invention;
[0024] Figure 3 A schematic diagram of an embodiment of a risk factor-based level calculation device in an embodiment of the present invention;
[0025] Figure 4 It is a schematic diagram of another embodiment of a level calculation device based on risk factors in an embodiment of the present invention;
[0026] Figure 5 The figure is a schematic diagram of an embodiment of a risk factor-based level calculation device in an embodiment of the present invention. DETAILED DESCRIPTION
[0027] The embodiments of the present invention provide a risk factor-based level calculation method, apparatus, device and storage medium for improving the accuracy of calculating correlation data, thereby improving the accuracy of matching target risk levels.
[0028] The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" or "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 necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.
[0029] For ease of understanding, the specific process of the embodiment of the present invention is described below. Figure 1 An embodiment of a method for calculating a level based on risk factors in an embodiment of the present invention includes:
[0030] 101. Obtain historical basic business data of the comprehensive business area and historical executed business data of the comprehensive business area, and perform structured processing on the historical basic business data and the historical executed business data to generate structured historical business data;
[0031] The server obtains the historical basic business data of the comprehensive business area and the historical execution business data of the comprehensive business area, and performs structured processing on the start basic business data and the historical execution business data to generate structured historical business data. It should be emphasized that in order to further ensure the privacy and security of the above structured historical business data, the above structured historical business data can also be stored in a node of a blockchain.
[0032] In this embodiment, the comprehensive business area refers to a business area that is fixed in a geographical location. There can be multiple comprehensive business areas. For ease of description, a comprehensive business area is used for explanation in this embodiment. In this embodiment, the historical basic business data refers to flood insurance underwriting data, wherein the flood insurance underwriting data includes insurance underwriting area data, insurance underwriting subject matter data, insurance time data, etc. In this embodiment, the historical execution business data refers to flood insurance claim data, wherein the flood insurance claim data includes claim time data and insurance claim detail data. The historical basic business data includes multiple sub-historical basic business data and the historical execution business data includes multiple sub-historical execution business data. The sub-historical basic business data corresponds to the sub-historical execution business data one by one. For example, the flood insurance underwriting data of subject matter A corresponds to the flood insurance claim data of subject matter A.
[0033] It is understandable that the execution subject of the present invention may be a risk factor-based level calculation device, or a terminal or a server, which is not limited here. The embodiment of the present invention is described by taking a server as the execution subject as an example.
[0034] 102. Generate target risk factors based on structured historical business data and preset risk factor models;
[0035] The server generates target risk factors based on structured historical business data and preset risk factor models.
[0036] In this embodiment, the preset risk factor model is a risk factor model trained in advance. During the training process, the risk factor model will perform random tree correction and automatic adjustment of factor weights on randomly selected training factors. The random tree correction mainly removes training factors whose weight parameters are less than the weight threshold, and then adjusts the factor weights of the removed training factors, thereby generating a risk factor model trained in advance, that is, a preset risk factor model. In this step, the structured historical business data is processed and input into the preset risk factor model to generate a target risk factor. In this embodiment, the target risk factor can be the insurance area, the type of insurance subject matter, the insurance policy time, the insurance premium scale, the type of insurance claim subject matter, the insurance claim time, and the insurance claim amount.
[0037] 103. Extract multiple target risk sub-factors and multiple target risk parent factors from structured historical business data based on the gray association algorithm and the target risk factor, and perform calculations based on the multiple target risk sub-factors and multiple target risk parent factors to generate a target correlation data set, where the target risk sub-factors and the target risk parent factors correspond one to one;
[0038] The server extracts multiple target risk sub-factors and multiple target risk parent factors from the structured historical business data based on the target risk factor, and performs calculations based on the multiple target risk sub-factors and multiple target risk parent factors based on the grayscale association algorithm to generate a target association data set. The target risk sub-factors and the target risk parent factors correspond one to one, and the target association data set includes multiple target association data.
[0039] It should be noted that the gray correlation algorithm is a quantitative method for analyzing the degree of correlation between various factors. Risk calculation using the gray correlation algorithm can reduce calculation errors and improve the accuracy of risk calculation.
[0040] 104. Obtain a risk query instruction, calculate the risk level according to the risk query business identifier in the risk query instruction in combination with the target relevance data set and the structured historical business data, and generate a target risk level.
[0041] The server obtains the risk query instruction, and then generates a target risk level according to the risk query business identifier in the risk query instruction in combination with the target relevance data set and the structured historical business data.
[0042] The query sub-business area or query business object indicated by the risk query instruction is located in the comprehensive business area. The risk query instruction is "the risk level of business object D in business sub-area C". The server determines the query business data in the structured historical business data in combination with the risk query instruction, and then matches the risk level based on the query business data with the reference target relevance data set to generate the target risk level.
[0043] In an embodiment of the present invention, historical basic business data and historical execution business data are structured into structured historical business data, and then a target risk factor and a correlation data set are generated based on a risk factor model and a grayscale correlation algorithm. Finally, a target risk level corresponding to the risk query instruction is generated according to the risk query instruction and the correlation data set. The structured historical business data saves computing resources of the server. In business usage scenarios, the target risk factor is determined in combination with the risk factor model, and then a correlation data set is generated based on the target risk factor, thereby improving the accuracy of calculating the correlation data and thus improving the accuracy of matching the target risk level.
[0044] See also Figure 2 Another embodiment of the risk factor-based level calculation method in the embodiment of the present invention includes:
[0045] 201. Obtain historical basic business data of the comprehensive business area and historical executed business data of the comprehensive business area, and perform structured processing on the historical basic business data and the historical executed business data to generate structured historical business data;
[0046] The server obtains the historical basic business data of the comprehensive business area and the historical execution business data of the comprehensive business area, and performs structured processing on the start basic business data and the historical execution business data to generate structured historical business data. It should be emphasized that in order to further ensure the privacy and security of the above structured historical business data, the above structured historical business data can also be stored in a node of a blockchain.
[0047] In this embodiment, the comprehensive business area refers to a geographically fixed business area. There can be multiple comprehensive business areas. For ease of description, this embodiment uses a comprehensive business area for illustration. In this embodiment, the historical basic business data refers to flood insurance underwriting data, where the flood insurance underwriting data includes insurance underwriting area data, insurance underwriting subject matter data, insurance time data, etc. In this embodiment, the historical execution business data refers to flood insurance claim data, where the flood insurance claim data includes claim time data and insurance claim details data.
[0048] Specifically, the server obtains the historical basic business data of the comprehensive business area and the historical execution business data of the comprehensive business area; the server reads the sub-business area code corresponding to each sub-business area from the historical basic business data and the historical execution business data to obtain multiple sub-business area codes, and reads the business object code corresponding to each business object from the historical basic business data and the historical execution business data to obtain multiple business object codes, and the comprehensive business area includes multiple sub-business areas and multiple business objects; the server performs structured processing on the historical basic business data and the historical execution business data based on the multiple sub-business area codes and the multiple business object codes to generate structured historical business data.
[0049] After the server obtains the historical basic business data and the historical execution business data from the database, it needs to perform structured processing on the data. In the process of structuring the data, it needs to combine the business coding to structure the data. First, the sub-business area code corresponding to each sub-business area is read from the historical basic business data and the historical execution business data to obtain multiple sub-business area codes. At the same time, the business object code corresponding to each business object is read from the historical basic business data and the historical execution business data to obtain multiple business object codes. It should be noted that the comprehensive area includes multiple sub-business areas and multiple business objects, each sub-business area has a corresponding sub-business area code, and each business object also has a corresponding business object code. The server performs structured processing on the historical basic business data and the historical execution business data based on multiple sub-business object codes and multiple business object codes to generate structured historical business data.
[0050] It should be noted that the historical basic data includes multiple historical basic sub-data, and the historical execution business data includes multiple historical execution business sub-data, wherein the historical basic sub-data, historical execution business sub-data, sub-business areas and business objects correspond one to one.
[0051] The server performs structured processing on historical basic business data and historical execution business data based on multiple sub-business area codes and multiple business object codes, and generates structured historical business data including:
[0052] The server obtains the cleanup instruction, and performs data cleanup on the historical basic business data and the historical execution business data in accordance with the cleanup instruction in combination with multiple sub-business area codes and multiple business object codes, to generate the historical basic business data after data cleansing and the historical execution business data after data cleansing; the server performs structured processing on the historical basic business data after data cleansing and the historical execution business data after data cleansing in combination with multiple sub-business area codes and multiple business object codes, to generate structured historical business data.
[0053] It should be noted that when performing structured processing on historical basic business data and historical execution business data, it is necessary to clean the historical basic business data and historical execution business data to filter out redundant business sub-data and save server computing resources in subsequent calculations.
[0054] In this embodiment, the coding format of the sub-business area code is "0001", and the coding format of the business object code is "a001". For example, the sub-business area code of the Jing'an District of Shanghai is 0001, and the business object code of the business object building A is a001. The server obtains a cleaning instruction, and the cleaning instruction is "clean the historical basic sub-data with a sub-business area code greater than 0199 and check the historical execution business sub-data". Then the service performs data cleaning on the historical basic business data and the historical execution business data based on the cleaning instruction combined with the sub-business area code and the business object code to generate the historical basic business data after data cleaning and the historical execution business data after data cleaning. Then, the historical basic business data after data cleaning and the historical execution business data after data cleaning are structured in the form of a relational database, mainly storing the historical basic business sub-data after data cleaning and the historical execution business sub-data in a one-to-one correspondence in the relational database, so that the historical basic business sub-data after data cleaning and the historical execution business sub-data form a data association relationship, and generate structured historical business data.
[0055] 202. Generate target risk factors based on structured historical business data and preset risk factor models;
[0056] The server generates target risk factors based on structured historical business data and preset risk factor models.
[0057] In this embodiment, the preset risk factor model is a risk factor model trained in advance based on a structured business training data set. The structured business training data set is data in a database, and can be directly extracted for model training. During the training process, the risk factor model will perform random tree correction and automatic adjustment of factor weights on the randomly selected training factors. The random tree correction mainly removes the training factors whose weight parameters are less than the weight threshold, and then adjusts the factor weights of the removed training factors, thereby generating a risk factor model trained in advance, that is, a preset risk factor model. The preset risk factor model can accurately select the target risk factor from the structured business data. In this step, the structured historical business data is processed and input into the preset risk factor model to generate the target risk factor. In this embodiment, the target risk factor can be the insurance area, the type of insurance subject matter, the insurance policy time, the insurance premium scale, the type of insurance claim subject matter, the insurance claim time and the insurance claim amount, etc.
[0058] Specifically, the server calls a preset random function to randomly select multiple initial risk factors from structured historical business data; the server inputs the multiple initial risk factors into a preset risk factor model to perform weight calculation, generates business weight parameters corresponding to each initial risk factor, and obtains multiple business weight parameters; the server determines the largest business weight parameter from the multiple business weight parameters as the target business weight parameter, and determines the initial risk factor corresponding to the target business weight parameter as the target risk factor.
[0059] The server calls a preset random function to randomly select multiple initial risk factors from the structured historical business data, for example, the initial risk factors are selected as insurance area, insurance subject type, insurance policy time, and insurance premium scale. Then the server inputs the insurance area, insurance subject type, insurance policy time, and insurance premium scale into the preset risk factor model for weight calculation, and generates a business weight parameter corresponding to the insurance area of 0.87, a business weight parameter corresponding to the insurance policy time of 0.73, a weight parameter corresponding to the insurance subject type of 0.78, and a weight parameter corresponding to the insurance premium scale of 0.96. The server determines the business weight parameter corresponding to the insurance premium scale of 0.96 as the target business weight parameter, and determines its corresponding initial risk factor, insurance premium scale, as the target risk factor.
[0060] 203. Extract multiple target risk sub-factors and multiple target risk parent factors from the structured historical business data based on the gray association algorithm and the target risk factor, and perform calculations based on the multiple target risk sub-factors and multiple target risk parent factors to generate a target correlation data set, where the target risk sub-factors and the target risk parent factors correspond one to one;
[0061] The server extracts multiple target risk sub-factors and multiple target risk parent factors from the structured historical business data based on the target risk factor, and performs calculations based on the multiple target risk sub-factors and multiple target risk parent factors based on the grayscale association algorithm to generate a target association data set. The target risk sub-factors and the target risk parent factors correspond one to one, and the target association data set includes multiple target association data.
[0062] It should be noted that the gray correlation algorithm is a quantitative method for analyzing the degree of correlation between various factors. Risk calculation using the gray correlation algorithm can reduce calculation errors and improve the accuracy of risk calculation.
[0063] Specifically, the server calls a preset child extraction function to extract sub-factors from the structured historical business data based on the target risk factor, and generates multiple target risk sub-factors; the server uses a preset parent extraction function to extract parent factors from the structured historical business data based on the target risk factor and each target risk sub-factor, and generates a target risk parent factor corresponding to each target risk sub-factor, and obtains multiple target risk parent factors; the server calculates the correlation between each target risk sub-factor and the corresponding target risk parent factor based on the grayscale correlation algorithm, generates initial correlation data corresponding to each target risk sub-factor, and obtains an initial correlation data set; the server sorts multiple initial correlation data in the initial correlation data set in order from large to small, and generates a target correlation data set, which includes multiple target correlation data.
[0064] It should be noted that the structured historical business data includes multiple structured historical business sub-data, so it is necessary to calculate the relevance of each structured historical business sub-data to generate the relevance data corresponding to each structured historical business sub-data, thereby obtaining the relevance data set. The server first calls the extract function to extract multiple target risk sub-factors from the structured historical business data based on the target risk factor. One target risk sub-factor corresponds to one structured historical business sub-data. The extract function is as follows:
[0065] V i =(u ij ),
[0066] where u ijis the j-th risk factor of the i-th disaster, that is, the risk factor corresponding to the target risk factor in a structured historical business sub-data.
[0067] The server then calls the parent function to extract multiple target risk parent factors from the structured historical business data based on the target risk factor. One target risk parent factor corresponds to one structured historical business sub-data and one target risk sub-factor. The parent function is as follows:
[0068] V0=(u 0j )=(1,2,…,m), that is, u 0j =1,(j=1,2,…,m),
[0069] Here, the value of m is a positive integer.
[0070] The server then calculates the difference sequence based on each target risk sub-factor and each target risk parent factor. The calculation process is as follows:
[0071] Δ 0j (j)=|V0(u 0j )-V i (u ij )|,
[0072] The calculation process of the difference sequence can be understood as calculating the absolute value of the difference between each target risk sub-factor and the corresponding target risk parent factor to generate the target factor difference;
[0073] The server then calculates the maximum difference and the minimum difference of the target factor corresponding to each target risk sub-factor. The calculation process is:
[0074]
[0075]
[0076] The server then calculates the risk correlation coefficient corresponding to each target risk sub-factor based on the target factor difference, the target factor maximum difference, and the target factor minimum difference. The calculation process is:
[0077]
[0078] The server finally calculates the risk correlation coefficient based on the grayscale correlation algorithm to generate target correlation data corresponding to each target risk sub-factor. The calculation process is:
[0079]
[0080] According to the above calculation method, each structured historical business sub-data is calculated to generate multiple target relevance data, thereby obtaining a target relevance data set.
[0081] 204. Obtain a risk query instruction, calculate the risk level according to the risk query business identifier in the risk query instruction in combination with the target correlation data set and the structured historical business data, and generate a target risk level.
[0082] The server obtains the risk query instruction, and then generates a target risk level according to the risk query business identifier in the risk query instruction in combination with the target relevance data set and the structured historical business data.
[0083] The risk query instruction is "the risk level of business object D in business sub-area C". The server determines the query business data in the structured historical business data based on the risk query instruction, and then matches the risk level with the target association data set based on the query business data to generate the target risk level.
[0084] Specifically, the server obtains a risk query instruction, parses the risk query instruction, generates a risk query business identifier, and performs factor query in structured historical business data based on the risk query business identifier to generate risk sub-factors to be queried; the server classifies the target correlation data set according to a preset classification rule to generate multiple risk levels; the server reads the target correlation data corresponding to the risk sub-factor to be queried in the target correlation data set, and matches the risk level based on the target correlation data to obtain the target risk level.
[0085] 205. Transmit the target risk level corresponding to the risk query instruction to the terminal corresponding to the risk query instruction.
[0086] After generating the target risk level based on the risk query instruction, the server transmits the target risk level to the terminal corresponding to the risk query instruction, thereby providing a basis for adjusting the insurance product and realizing the maximum value of the insurance product.
[0087] In an embodiment of the present invention, historical basic business data and historical execution business data are structured into structured historical business data, and then a target risk factor and a correlation data set are generated based on a risk factor model and a grayscale correlation algorithm. Finally, a target risk level corresponding to the risk query instruction is generated according to the risk query instruction and the correlation data set. The structured historical business data saves computing resources of the server. In business usage scenarios, the target risk factor is determined in combination with the risk factor model, and then a correlation data set is generated based on the target risk factor, thereby improving the accuracy of calculating the correlation data and thus improving the accuracy of matching the target risk level.
[0088] The above describes the level calculation method based on risk factors in the embodiment of the present invention. The following describes the level calculation device based on risk factors in the embodiment of the present invention. Figure 3 In one embodiment of the present invention, a risk factor-based level calculation device includes:
[0089] The acquisition module 301 is used to acquire the historical basic business data of the integrated business area and the historical executed business data of the integrated business area, and perform structured processing on the historical basic business data and the historical executed business data to generate structured historical business data;
[0090] A risk factor generation module 302, configured to generate a target risk factor according to the structured historical business data and a preset risk factor model;
[0091] A calculation module 303 is used to extract multiple target risk sub-factors and multiple target risk parent factors from the structured historical business data based on the gray association algorithm and the target risk factor, and perform calculations based on the multiple target risk sub-factors and the multiple target risk parent factors to generate a target correlation data set, where the target risk sub-factors and the target risk parent factors correspond one to one;
[0092] The query module 304 is used to obtain a risk query instruction, calculate the risk level according to the risk query business identifier in the risk query instruction in combination with the target relevance data set and the structured historical business data, and generate a target risk level.
[0093] In an embodiment of the present invention, historical basic business data and historical execution business data are structured into structured historical business data, and then a target risk factor and a correlation data set are generated based on a risk factor model and a grayscale correlation algorithm. Finally, a target risk level corresponding to the risk query instruction is generated according to the risk query instruction and the correlation data set. The structured historical business data saves computing resources of the server. In business usage scenarios, the target risk factor is determined in combination with the risk factor model, and then a correlation data set is generated based on the target risk factor, thereby improving the accuracy of calculating the correlation data and thus improving the accuracy of matching the target risk level.
[0094] See also Figure 4 Another embodiment of the risk factor-based level calculation device in the embodiment of the present invention includes:
[0095] The acquisition module 301 is used to acquire the historical basic business data of the integrated business area and the historical executed business data of the integrated business area, and perform structured processing on the historical basic business data and the historical executed business data to generate structured historical business data;
[0096] A risk factor generation module 302, configured to generate a target risk factor according to the structured historical business data and a preset risk factor model;
[0097] A calculation module 303 is used to extract multiple target risk sub-factors and multiple target risk parent factors from the structured historical business data based on the gray association algorithm and the target risk factor, and perform calculations based on the multiple target risk sub-factors and the multiple target risk parent factors to generate a target correlation data set, where the target risk sub-factors and the target risk parent factors correspond one to one;
[0098] The query module 304 is used to obtain a risk query instruction, calculate the risk level according to the risk query business identifier in the risk query instruction in combination with the target relevance data set and the structured historical business data, and generate a target risk level.
[0099] Optionally, the acquisition module 301 includes:
[0100] An acquisition unit 3011 is used to acquire historical basic business data of the integrated business area and historical executed business data of the integrated business area;
[0101] The reading unit 3012 is used to read the sub-business area code corresponding to each sub-business area from the historical basic business data and the historical execution business data to obtain a plurality of sub-business area codes, and read the business object code corresponding to each business object from the historical basic business data and the historical execution business data to obtain a plurality of business object codes, wherein the comprehensive business area includes a plurality of sub-business areas and a plurality of business objects;
[0102] The structuring unit 3013 is used to perform structural processing on the historical basic business data and the historical execution business data based on the multiple sub-business area codes and the multiple business object codes to generate structured historical business data.
[0103] Optionally, the structuring unit 3013 may also be specifically used for:
[0104] Obtaining a cleanup instruction, and performing data cleanup on the historical basic business data and the historical executed business data in accordance with the cleanup instruction in combination with the multiple sub-business area codes and the multiple business object codes, to generate the cleaned historical basic business data and the cleaned historical executed business data;
[0105] The historical basic business data after data cleaning and the historical execution business data after data cleaning are structured in combination with the multiple sub-business area codes and the multiple business object codes to generate structured historical business data.
[0106] Optionally, the risk factor generating module 302 may also be specifically used for:
[0107] Calling a preset random function to randomly select a plurality of initial risk factors from the structured historical business data;
[0108] Inputting the multiple initial risk factors into a preset risk factor model to perform weight calculation, generating a business weight parameter corresponding to each initial risk factor, and obtaining multiple business weight parameters;
[0109] The largest business weight parameter is determined from the multiple business weight parameters as the target business weight parameter, and the initial risk factor corresponding to the target business weight parameter is determined as the target risk factor.
[0110] Optionally, the calculation module 303 may also be specifically used for:
[0111] Calling a preset sub-factor extraction function to extract sub-factors from the structured historical business data based on the target risk factor to generate a plurality of target risk sub-factors;
[0112] Calling a preset parent function to extract a parent factor from the structured historical business data based on the target risk factor and each target risk sub-factor, generating a target risk parent factor corresponding to each target risk sub-factor, and obtaining a plurality of target risk parent factors;
[0113] Based on the gray correlation algorithm, the correlation between each target risk sub-factor and the corresponding target risk parent factor is calculated to generate initial correlation data corresponding to each target risk sub-factor, thereby obtaining an initial correlation data set;
[0114] The multiple initial relevance data in the initial relevance data set are sorted in order from large to small to generate a target relevance data set, wherein the target relevance data set includes multiple target relevance data.
[0115] Optionally, the query module 304 may also be specifically used for:
[0116] Obtaining a risk query instruction, parsing the risk query instruction, generating a risk query service identifier, and performing a factor query in the structured historical service data based on the risk query service identifier to generate a risk sub-factor to be queried;
[0117] Classify the target relevance data set according to a preset classification rule to generate multiple risk levels;
[0118] The target relevance data corresponding to the risk sub-factor to be queried is read from the target relevance data set, and the risk level is matched based on the target relevance data to obtain the target risk level.
[0119] Optionally, the risk factor-based level calculation device further includes:
[0120] The transmission module 305 is used to transmit the target risk level corresponding to the risk query instruction to the terminal corresponding to the risk query instruction.
[0121] In an embodiment of the present invention, historical basic business data and historical execution business data are structured into structured historical business data, and then a target risk factor and a correlation data set are generated based on a risk factor model and a grayscale correlation algorithm. Finally, a target risk level corresponding to the risk query instruction is generated according to the risk query instruction and the correlation data set. The structured historical business data saves computing resources of the server. In business usage scenarios, the target risk factor is determined in combination with the risk factor model, and then a correlation data set is generated based on the target risk factor, thereby improving the accuracy of calculating the correlation data and thus improving the accuracy of matching the target risk level.
[0122] above Figure 3 and Figure 4 The level calculation device based on risk factors in the embodiment of the present invention is described in detail from the perspective of modular functional entities. The level calculation device based on risk factors in the embodiment of the present invention is described in detail from the perspective of hardware processing.
[0123] Figure 5 1 is a schematic diagram of a structure of a risk factor-based ranking calculation device provided by an embodiment of the present invention. The risk factor-based ranking calculation device 500 may have relatively large differences due to different configurations or performances, and may include one or more processors (central processing units, CPU) 510 (for example, one or more processors) and a memory 520, and one or more storage media 530 (for example, one or more mass storage devices) storing application programs 533 or data 532. Among them, the memory 520 and the storage medium 530 may be temporary storage or permanent storage. The program stored in the storage medium 530 may include one or more modules (not shown in the figure), and each module may include a series of instruction operations in the risk factor-based ranking calculation device 500. Furthermore, the processor 510 may be configured to communicate with the storage medium 530 to execute a series of instruction operations in the storage medium 530 on the risk factor-based ranking calculation device 500.
[0124] The risk factor-based ranking computing device 500 may also include one or more power supplies 540, one or more wired or wireless network interfaces 550, one or more input and output interfaces 560, and / or one or more operating systems 531, such as Windows Serve, Mac OS X, Unix, Linux, FreeBSD, etc. It will be appreciated by those skilled in the art that Figure 5 The structure of the risk factor-based rank calculation device shown does not constitute a limitation on the risk factor-based rank calculation device, and may include more or less components than shown in the figure, or combine some components, or arrange the components differently.
[0125] The present invention also provides a risk factor-based grade calculation device, the computer device comprising a memory and a processor, the memory storing computer-readable instructions, and when the computer-readable instructions are executed by the processor, the processor executes the steps of the risk factor-based grade calculation method in the above-mentioned embodiments.
[0126] The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein instructions are stored in the computer, and when the instructions are executed on a computer, the computer executes the steps of the risk factor-based level calculation method.
[0127] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.
[0128] The blockchain referred to in this invention is a new application mode of computer technologies such as distributed data storage, peer-to-peer transmission, consensus mechanism, encryption algorithm, etc. Blockchain is essentially a decentralized database, a string of data blocks generated by cryptographic methods. Each data block contains a batch of network transaction information, which is used to verify the validity of its information (anti-counterfeiting) and generate the next block. Blockchain can include the underlying blockchain platform, platform product service layer, and application service layer.
[0129] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program codes.
[0130] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A method for calculating a level based on risk factors, characterized in that: The risk factor-based grade calculation method includes: Acquire historical basic business data of the integrated business area and historical executed business data of the integrated business area, and perform structured processing on the historical basic business data and the historical executed business data to generate structured historical business data; Calling a preset random function to randomly select multiple initial risk factors from the structured historical business data; inputting the multiple initial risk factors into a preset risk factor model to perform weight calculation, generating a business weight parameter corresponding to each initial risk factor, and obtaining multiple business weight parameters; determining the largest business weight parameter from the multiple business weight parameters as the target business weight parameter, and determining the initial risk factor corresponding to the target business weight parameter as the target risk factor; Calling a preset child extraction function to extract sub-factors from the structured historical business data based on the target risk factor, and generating multiple target risk sub-factors; calling a preset parent extraction function to extract parent factors from the structured historical business data based on the target risk factor and each target risk sub-factor, and generating a target risk parent factor corresponding to each target risk sub-factor, and obtaining multiple target risk parent factors; calculating the correlation between each target risk sub-factor and the corresponding target risk parent factor based on the grayscale correlation algorithm, generating initial correlation data corresponding to each target risk sub-factor, and obtaining an initial correlation data set; sorting multiple initial correlation data in the initial correlation data set in order from large to small, and generating a target correlation data set, wherein the target correlation data set includes multiple target correlation data; Obtain a risk query instruction, parse the risk query instruction, generate a risk query business identifier, and perform factor query in the structured historical business data based on the risk query business identifier to generate risk sub-factors to be queried; grade the target association data set according to a preset grade classification rule to generate multiple risk levels; read the target association data set corresponding to the risk sub-factor to be queried, and match the risk level based on the target association data to obtain the target risk level.
2. The risk factor-based level calculation method according to claim 1, characterized in that: The acquiring of historical basic business data of the integrated business area and historical executed business data of the integrated business area, and performing structured processing on the historical basic business data and the historical executed business data to generate structured historical business data includes: Obtain historical basic business data of the comprehensive business area and historical execution business data of the comprehensive business area; Reading a sub-business area code corresponding to each sub-business area from the historical basic business data and the historical execution business data to obtain a plurality of sub-business area codes, and reading a business object code corresponding to each business object from the historical basic business data and the historical execution business data to obtain a plurality of business object codes, wherein the comprehensive business area includes a plurality of sub-business areas and a plurality of business objects; The historical basic business data and the historical execution business data are structured based on the multiple sub-business area codes and the multiple business object codes to generate structured historical business data.
3. The risk factor-based level calculation method according to claim 2, characterized in that: The step of performing structural processing on the historical basic business data and the historical execution business data based on the multiple sub-business area codes and the multiple business object codes to generate structured historical business data includes: Obtaining a cleanup instruction, and performing data cleanup on the historical basic business data and the historical executed business data in accordance with the cleanup instruction in combination with the multiple sub-business area codes and the multiple business object codes, to generate the cleaned historical basic business data and the cleaned historical executed business data; The historical basic business data after data cleaning and the historical execution business data after data cleaning are structured in combination with the multiple sub-business area codes and the multiple business object codes to generate structured historical business data.
4. The method for calculating a level based on risk factors according to any one of claims 1 to 3, characterized in that: In the acquiring risk query instruction, the risk level calculation is performed according to the risk query business identifier in the risk query instruction in combination with the target association data set and the structured historical business data to generate a target risk level, and the level calculation method based on risk factors further includes: The target risk level corresponding to the risk query instruction is transmitted to the terminal corresponding to the risk query instruction.
5. A risk factor-based level calculation device, characterized in that: The risk factor-based level calculation device includes: An acquisition module, used to acquire historical basic business data of the integrated business area and historical executed business data of the integrated business area, and perform structured processing on the historical basic business data and the historical executed business data to generate structured historical business data; A risk factor generation module is used to call a preset random function to randomly select multiple initial risk factors from the structured historical business data; input the multiple initial risk factors into a preset risk factor model to perform weight calculation, generate a business weight parameter corresponding to each initial risk factor, and obtain multiple business weight parameters; determine the largest business weight parameter from the multiple business weight parameters as the target business weight parameter, and determine the initial risk factor corresponding to the target business weight parameter as the target risk factor; A calculation module, used for calling a preset child extraction function to extract sub-factors from the structured historical business data based on the target risk factor, and generating multiple target risk sub-factors; calling a preset parent extraction function to extract parent factors from the structured historical business data based on the target risk factor and each target risk sub-factor, and generating a target risk parent factor corresponding to each target risk sub-factor, and obtaining multiple target risk parent factors; calculating the correlation between each target risk sub-factor and the corresponding target risk parent factor based on a gray association algorithm, generating initial correlation data corresponding to each target risk sub-factor, and obtaining an initial correlation data set; sorting multiple initial correlation data in the initial correlation data set in order from large to small, and generating a target correlation data set, wherein the target correlation data set includes multiple target correlation data; A query module is used to obtain a risk query instruction, parse the risk query instruction, generate a risk query business identifier, and perform factor query in the structured historical business data based on the risk query business identifier to generate risk sub-factors to be queried; classify the target association data set according to a preset grade classification rule to generate multiple risk levels; read the target association data set corresponding to the risk sub-factor to be queried, and match the risk level based on the target association data to obtain the target risk level.
6. A risk factor-based level calculation device, characterized in that: The risk factor-based level calculation device includes: a memory and at least one processor, wherein instructions are stored in the memory; The at least one processor calls the instructions in the memory to enable the risk factor-based grade calculation device to execute the risk factor-based grade calculation method according to any one of claims 1-4.
7. A computer-readable storage medium having instructions stored thereon, characterized in that: When the instructions are executed by a processor, a level calculation method based on risk factors as described in any one of claims 1 to 4 is implemented.
