WO2022126975A1 - 客户信息校验方法、装置、计算机设备及存储介质 - Google Patents

客户信息校验方法、装置、计算机设备及存储介质 Download PDF

Info

Publication number
WO2022126975A1
WO2022126975A1 PCT/CN2021/090587 CN2021090587W WO2022126975A1 WO 2022126975 A1 WO2022126975 A1 WO 2022126975A1 CN 2021090587 W CN2021090587 W CN 2021090587W WO 2022126975 A1 WO2022126975 A1 WO 2022126975A1
Authority
WO
WIPO (PCT)
Prior art keywords
verification
information
sample
risk
customer
Prior art date
Application number
PCT/CN2021/090587
Other languages
English (en)
French (fr)
Inventor
李春平
Original Assignee
平安科技（深圳）有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 平安科技（深圳）有限公司 filed Critical 平安科技（深圳）有限公司
Publication of WO2022126975A1 publication Critical patent/WO2022126975A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0635—Risk analysis of enterprise or organisation activities
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/903—Querying
    • G06F16/90335—Query processing

Definitions

  • the present application relates to the field of artificial intelligence technology, and belongs to the application scenario of intelligent risk verification of customer information in a smart city, and in particular relates to a customer information verification method, device, computer equipment and storage medium.
  • the enterprise In the process of verification business processing, the enterprise will obtain customer information and verify it, so as to reduce the risk of business processing through the verification process. If the customer information meets the corresponding processing conditions, the business will be processed. Otherwise, there is a problem in the customer information. , the enterprise needs to deal with it according to the severity of the problem in the customer information.
  • the traditional technical methods all use the judgment method to judge and verify the customer information, so as to obtain the verification result of the customer information.
  • the customer information contains the information of association, mutual exclusion, etc., for financial enterprises to judge the customer information.
  • the logic is even more complicated, which makes it take a long time to judge customer information using judgment statements. Therefore, the inventor found that it is difficult to verify the massive concurrent customer information in real time and efficiently with traditional technical methods. It takes a long time to complete the verification of customer information, which affects the timeliness of the company's follow-up business processing based on customer information. Therefore, the prior art method has the problem that the customer information cannot be verified in real time and efficiently.
  • the embodiments of the present application provide a method, device, computer equipment and storage medium for verifying customer information, which aim to solve the problem of inability to verify customer information in real time and efficiently in the prior art methods.
  • an embodiment of the present application provides a method for verifying customer information, which includes:
  • sample customer information is randomly selected from the pre-stored historical customer information table
  • risk verification is performed on the newly added customer information sent by the client in real time, and a new client risk verification result is obtained and fed back to the client.
  • an embodiment of the present application provides a customer information verification device, which includes:
  • the sample customer information acquisition unit is used to randomly extract sample customer information from the pre-stored historical customer information table if the risk verification rule input by the administrator is received;
  • a sample verification result obtaining unit configured to perform risk verification on the sample customer information according to the risk verification rule to obtain a sample verification result for each of the sample customer information
  • a sample customer quantitative information acquisition unit configured to quantify the sample customer information according to a preset information quantification rule to obtain corresponding sample customer quantitative information
  • a risk verification model update unit configured to iteratively update the preset risk verification model according to the pre-stored model update rule, the sample customer quantitative information and the sample verification result, to obtain an updated risk verification model
  • a risk verification unit configured to perform risk verification on the newly added customer information sent by the client in real time according to the information quantification rules and the updated risk verification model, to obtain the new customer risk verification result and feedback it to the client.
  • an embodiment of the present application further provides a computer device, which includes a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor executes the computer In the program, the client information verification method described in the first aspect above is implemented.
  • an embodiment of the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when executed by a processor, the computer program causes the processor to execute the above-mentioned first step.
  • Embodiments of the present application provide a client information verification method, device, computer equipment, and storage medium. Randomly select sample customer information from the historical customer information table, perform risk verification on the sample customer information according to the risk verification rules to obtain the sample verification result, and quantify the sample customer information to obtain the sample customer quantitative information.
  • the quantitative customer information updates the risk verification model iteratively, and uses the updated risk verification model to perform risk verification on the new customer information sent by the client in real time to obtain the new customer risk verification result.
  • FIG. 1 is a schematic flowchart of a client information verification method provided in an embodiment of the present application.
  • FIG. 2 is a schematic diagram of an application scenario of the method for verifying customer information provided by an embodiment of the present application
  • FIG. 3 is a schematic diagram of a sub-flow of a method for verifying client information provided by an embodiment of the present application
  • FIG. 4 is a schematic diagram of another sub-flow of the client information verification method provided by the embodiment of the present application.
  • FIG. 5 is a schematic diagram of another sub-flow of the client information verification method provided by the embodiment of the present application.
  • FIG. 6 is a schematic diagram of another sub-flow of the client information verification method provided by the embodiment of the present application.
  • FIG. 7 is a schematic diagram of another sub-flow of the customer information verification method provided by the embodiment of the present application.
  • FIG. 8 is another schematic flowchart of the customer information verification method provided by the embodiment of the present application.
  • FIG. 9 is a schematic block diagram of a client information verification device provided by an embodiment of the present application.
  • FIG. 10 is a schematic block diagram of a computer device provided by an embodiment of the present application.
  • FIG. 1 is a schematic flowchart of a method for verifying client information provided by an embodiment of the present application
  • FIG. 2 is a schematic diagram of an application scenario of the method for verifying client information provided by an embodiment of the present application.
  • the method is applied in the management server 10, the method is executed by the application software installed in the management server 10, the management server 10 is connected to at least one client 20 through a network to realize the transmission of data information, and the management server 10 is used for Execute the client information verification method to realize the server side of intelligent risk verification for client information.
  • the management server 10 can be a server established in the enterprise, and the user of the management server 10 is the administrator of the enterprise; the client 20 is the A terminal device that establishes a network connection with the management server 10 for data information transmission, such as a desktop computer, a notebook computer, a tablet computer or a mobile phone, etc., the user of the client terminal 20 is a client. As shown in FIG. 1, the method includes steps S110-S150.
  • sample customer information is randomly selected from the pre-stored historical customer information table.
  • sample customer information is randomly selected from the pre-stored historical customer information table.
  • the administrator can configure risk verification rules to the management server, where the input risk verification rules can be newly configured rules, or rules obtained by modifying old risk verification rules. That is, the rule information for risk verification of customer information.
  • the risk verification rule contains multiple rules that can be specifically represented by logical operators, and the customer information can be verified based on the logical operators in the risk verification rule. If the input risk verification rule is received, part of the customer information is randomly selected from the historical customer information table as sample customer information.
  • the historical customer information table is the information table configured in the management server to store historical customer information. When customers handle business, they will send customer information to the management server through the client, and the management server can receive the customer information.
  • customer information can include customer name, age, gender, ID number, mobile phone number, occupation, income, hobbies, housing information, private car information, address, marital status, birth information and credit information Information related to customers such as default information.
  • sample customer information that matches the proportional value can be randomly selected from the pre-stored historical customer information table according to the preset ratio value; the historical customer information table stores a large amount of customer information. Then, part of the customer information can be randomly selected as sample customer information according to the preset ratio value.
  • the historical customer information table contains 10,000 pieces of customer information, and the preset ratio value is 0.1, correspondingly, 1,000 pieces of customer information are extracted from the historical customer information table as sample customer information.
  • the sample customer information can be subjected to risk verification according to the risk verification rules to obtain the corresponding sample verification results, and the sample verification results can be verified or failed, wherein the risk verification rules include format verification rules, Correlation check rules and matching check rules.
  • the step S120 includes sub-steps: S121 , S122 , S123 , S124 and S125 .
  • Format verification can be performed on multiple pieces of information contained in the sample customer information according to the format verification rules.
  • the format verification rules include specific rules for verifying the character length and character type of each item of information.
  • Format verification The rule contains the verification format corresponding to multiple pieces of information. If the character length or character type of a certain piece of information in the sample customer information does not meet the verification format corresponding to the information, the format verification result of the sample customer information is: Failed; if the character length and character type of each item of information in the sample customer information meet the verification format corresponding to the information, the format verification result is passed.
  • the verification format of the ID number information includes: the character length is 18 characters, the character type of the first 17 characters must be numbers, and the character type of the last character is a number or letter.
  • the verification format of the certificate is used to verify one piece of information corresponding to the ID card in each sample customer information.
  • the correlation between multiple pieces of information in the sample customer information can be verified according to the correlation verification rules.
  • the correlation verification rules include specific rules for correlation verification of multiple pairs of correlation information in the customer information.
  • Each association information pair contains two pieces of information, and the association features of the two pieces of information corresponding to one association information pair can be obtained according to the association verification rules, and whether the association features are consistent, if the association features of the two pieces of information are consistent, it indicates that The two pieces of information of the associated information pair are correlated accordingly, otherwise it indicates that there is no correlation between the two pieces of information of the correlated information pair. If there is no correlation between the two pieces of information corresponding to a certain sample customer information and a certain pair of related information, the correlation verification result of the sample customer information is not passed; If the item information is associated, the association verification result is passed.
  • two pieces of information corresponding to a pair of associated information in a sample customer information are "mobile phone number: 138XXXXXXX” and "address: District E, D street, district C, city B, province A”, and the obtained “mobile phone number: 138XXXXXXX” ” is associated with city A, province B, and address associated with city A, province B, and the associated features of the two pieces of information are consistent, then the two pieces of information are correlated.
  • the matching verification rule includes a preset scope corresponding to each item of information, an item of information.
  • the corresponding preset range can be a set that defines the scope of the item of information, and it can be determined whether each item of information in the sample customer information matches the preset range according to the matching check rule. If a piece of information does not match the preset range, the matching verification result of the sample information will be failed; if each piece of information in a sample customer information matches the preset scope, the matching verification result will be passed .
  • the credit default information in a sample customer information is 5 times of default, and the preset range corresponding to the credit default information in the matching check rule is [0,3], then the credit default number of the sample customer information If it does not match the corresponding preset range, the matching verification result of the sample customer information is not passed.
  • the sample verification result of the sample customer information is passed; if any one of the format verification result, the association verification result and the matching verification result is not passed, then the sample client information sample is obtained. The verification result is failed.
  • the sample customer information is quantified according to a preset information quantification rule to obtain corresponding sample customer quantitative information.
  • Information quantification rules are specific rules for quantifying each sample customer information. According to the information quantification rules, each item of information contained in each sample customer information can be quantified to obtain the corresponding sample customer quantification information. Sample customer quantification The information can be used to quantify the information of each sample customer.
  • the information quantification rule includes multiple quantitative items. Each quantitative item can convert one piece of The quantified value constitutes the sample customer quantitative information that constitutes the sample customer information.
  • step S130 includes sub-steps S131 and S132.
  • the information quantification rule may include multiple quantification items, and item attribute information corresponding to each sample customer information and each quantification item may be sequentially obtained according to the multiple quantification items.
  • S132 Perform quantitative processing on item attribute information corresponding to each of the sample customer information according to the item rule of each of the quantified items, to obtain sample customer quantitative information of each sample customer information.
  • the item rule can quantify the item attribute information that matches the quantitative item.
  • the item rule of each quantitative item can convert an item attribute information into a quantitative value for representation, and a sample customer quantitative information can be represented as a A multi-dimensional feature vector, that is, a feature vector of one dimension in the sample customer quantization information corresponding to each item attribute information, and the range of quantization values obtained by quantizing the item attribute information corresponding to each quantization item is [0, 1 ].
  • the item attribute information belongs to a preset feature attribute, and if it belongs to a feature attribute, the item attribute information is directly converted into a corresponding feature attribute value, and the feature attribute includes ID number, mobile phone number, etc.; If the information does not belong to the feature attribute, it can be judged whether the item attribute information is a numerical value. If the item attribute information is a numerical value, the item rule matching the item attribute information is the activation function and a corresponding intermediate value, which can be calculated by the activation function. Obtain the quantitative value of the item attribute information; if the item attribute information is not a numerical value, the item rule matching the item attribute information contains multiple keywords and the value corresponding to each keyword, and obtains the item attribute in the item rule. The value corresponding to a keyword whose information matches is used as the quantitative value of the attribute information of the item.
  • the item attribute information is a feature attribute
  • the item attribute information is converted into a corresponding decimal for representation. If the ID number in a sample customer information is 210101XXXXXXXXXXX, the corresponding characteristic attribute value is 0.210101XXXXXXXXXXX.
  • the corresponding item rule is an activation function and an intermediate value. Get the corresponding quantized value.
  • the activation function in the item rule of a quantitative item can be expressed as: Among them, x is an item attribute information corresponding to a quantified item, and v is an intermediate value included in the item rule.
  • Marital status of the information quantification rule The item rule corresponding to the quantitative item contains three keywords: “married”, “divorced”, and “unmarried”, and the value corresponding to "married” is "1", and "divorced” The corresponding value is "0.3", and the value corresponding to "unmarried” is "0". If the marital status of a sample customer information is unmarried, the corresponding quantitative value is "0".
  • S140 Iteratively update the preset risk verification model according to the pre-stored model update rule, the sample customer quantitative information and the sample verification result, to obtain an updated risk verification model.
  • the preset risk verification model is iteratively updated according to the pre-stored model update rule, the sample customer quantitative information and the sample verification result, to obtain an updated risk verification model.
  • the model update rule is the rule for training and updating the parameter values in the risk verification model, and the model update rule includes the loss value calculation formula and the gradient calculation formula.
  • the risk verification model is an intelligent verification model constructed based on a neural network.
  • the risk verification model consists of an input layer, multiple intermediate layers and an output layer.
  • the number of input nodes contained in the input layer is equal to the number of dimensions of the quantitative information of sample customers, then each quantitative value in the quantitative information of sample customers corresponds to an input node, and the quantitative information of multiple sample customers is input into the risk verification in turn
  • the output result can be obtained from its output layer, and the loss value can be obtained by calculating the output result and the sample verification result corresponding to the quantitative information of the sample customer according to the model update rules, and then calculating the risk verification according to the loss value.
  • the updated value of each parameter value in the model can iteratively update the risk verification model.
  • the output result is the output node value of the output node
  • each quantized input information corresponds to two output node values
  • the output node value is the matching degree between the sample customer quantization information and the corresponding output node
  • the first output node value In order to verify the matching degree that passes, the second output node value is the matching degree that does not pass the verification.
  • the output node value can be represented by a decimal, and the value range is [0, 1].
  • step S140 includes sub-steps S141 , S142 , S143 , S144 , S145 and S146 .
  • the output node value corresponding to a sample customer quantitative information is obtained according to the risk verification model as the corresponding output result, and the output result includes the output node values corresponding to the two output nodes of the risk verification model respectively.
  • the sample verification result and output result can be calculated by the loss value calculation formula, and the loss value corresponding to the quantitative information of the sample customer can be obtained.
  • the loss value can be used to quantify the difference between the output result and the sample verification result.
  • the loss function can be expressed as Among them, R is the quantified value of the result corresponding to the sample verification result, the quantified value of the corresponding result of the sample verification result is “1”, the quantified value of the corresponding result of the sample verification result is “fail”, and r 1 is The output node value of the corresponding output node that passes the verification, r 1 is the output node value of the corresponding output node that does not pass the verification, and f(r) is the calculated loss value.
  • the updated value of each parameter in the risk verification model can be calculated according to the gradient calculation formula to update the original parameter value of the parameter. Specifically, a parameter in the risk verification model is calculated from the quantitative information of a sample customer. The calculated value is input into the gradient calculation formula, and combined with the above loss value, the updated value corresponding to the parameter can be calculated. This calculation process is also the gradient descent calculation.
  • the gradient calculation formula can be expressed as:
  • S144 Determine whether each of the sample customer quantitative information has iteratively updated the risk verification model; S145. If each of the sample customer quantitative information has iteratively updated the risk verification model, update The risk verification model is determined to be the updated risk verification model; S146, if the risk verification model is not iteratively updated for each of the sample customer quantitative information, obtain the next sample customer quantitative information and Return to and execute the step of obtaining an output result of the sample customer quantitative information according to the risk verification model.
  • the parameter values of all parameters in the risk verification model can be updated once, that is, a training update of the risk verification model is completed; a sample customer can be obtained. Quantify the information and repeat the above update process to iteratively update the risk verification model, until all sample customer quantitative information is used to iteratively update the risk verification model.
  • risk verification is performed on the newly added customer information sent by the client in real time, and a new client risk verification result is obtained and fed back to the client.
  • the management server can receive the new customer information sent by the client in real time, obtain the newly added customer quantitative information of the newly added customer information through the information quantification rules, and obtain the output result of the newly added customer quantitative information through the updated risk verification model. Further obtain the new customer risk verification results. If the risk verification rule input by the administrator is received again, the method in step S110 is executed back.
  • step S150 includes sub-steps S151 , S152 and S153 .
  • the newly added customer information can be quantified according to the information quantification rule, and the specific method of performing the quantization processing is the same as the specific method of performing the quantization processing on the sample customer quantified information, which will not be repeated here.
  • the risk level matching rule is a specific rule used to obtain the risk level matching the output result. The higher the risk level, the greater the business risk of the newly added customer information.
  • the risk level matching rule includes a normalization function and a risk level interval.
  • step S153 includes sub-steps S1531 and S1532.
  • S1531. Calculate and obtain a verification score according to the normalization function and the two output node values in the output result; S1532, acquire a level in the risk level interval that matches the verification score as a level that matches the verification score.
  • the output result matches the risk level.
  • the normalization function can be expressed as: Among them, r 1 is the output node value of the corresponding output node that passes the verification, r 1 is the output node value of the corresponding output node that does not pass the verification, and D is the calculated verification score.
  • the risk level interval includes a score interval corresponding to each level of multiple levels, obtains a score interval in the risk level interval that matches the verification score, and uses the level of the score interval as the risk level that matches the output result. .
  • the management server can carry out follow-up business processing for the new customer information whose risk verification result is no risk.
  • the management server can also add a risk label matching the risk level to the new customer information of different risk levels, and feed back the new customer information with the added risk label to the The client can modify it. If the risk verification result of the new customer is high risk, the corresponding prompt message can also be sent to the employee terminal of the enterprise to remind the corresponding employee of the enterprise to pay attention to risk prevention and control.
  • the verification score corresponding to an output result is 80.99
  • the verification score matches the score interval [75,90]
  • the level of "low risk" in this score interval is obtained as the risk matching the output result grade.
  • step S160 is further included after step S150 .
  • the blockchain referred to in this example is a new application mode of computer technology such as distributed data storage, point-to-point transmission, consensus mechanism, and encryption algorithm.
  • Blockchain essentially a decentralized database, is a series of data blocks associated with cryptographic methods. Each data block contains a batch of network transaction information to verify its Validity of information (anti-counterfeiting) and generation of the next block.
  • the blockchain can include the underlying platform of the blockchain, the platform product service layer, and the application service layer.
  • step S150 a step is further included: adding the newly added customer information to the historical customer information table for storage.
  • the newly added customer information can be added to the historical customer information table, and the newly added customer information added to the historical customer information table can be used as sample customer information for updating the risk verification model again.
  • the technical methods in this application can be applied to smart government affairs/smart city management/smart community/smart security/smart logistics/smart medical care/smart education/smart environmental protection/smart transportation and other application scenarios including intelligent risk verification of customer information, So as to promote the construction of smart cities.
  • sample customer information is randomly selected from the historical customer information table, and the sample customer information is subjected to risk verification according to the risk verification rule to obtain a sample verification result.
  • Quantify the sample customer quantitative information iteratively update the risk verification model according to the sample verification results and sample customer quantitative information, and use the updated risk verification model to perform risk verification on the new customer information sent by the client in real time. Get the new customer risk verification result.
  • the embodiment of the present application further provides a client information verification apparatus, and the client information verification apparatus is configured to execute any of the foregoing client information verification methods.
  • FIG. 9 is a schematic block diagram of an apparatus for verifying customer information provided by an embodiment of the present application.
  • the client information verification device may be arranged in the management server 10 .
  • the client information verification apparatus 100 includes a sample client information acquisition unit 110 , a sample verification result acquisition unit 120 , a sample client quantitative information acquisition unit 130 , a risk verification model update unit 140 and a risk verification unit 150 .
  • the sample customer information acquisition unit 110 is used for the sample customer information acquisition unit, and is used for randomly extracting sample customer information from the pre-stored historical customer information table if the risk verification rule input by the administrator is received.
  • the sample verification result obtaining unit 120 is configured to perform risk verification on the sample customer information according to the risk verification rule to obtain a sample verification result of each of the sample customer information.
  • the sample verification result obtaining unit 120 includes subunits: a format verification unit, an association verification unit, a match verification unit, a first judgment unit, and a second judgment unit.
  • a format verification unit configured to perform format verification on each item of information in each of the sample client information according to the format verification rule to obtain a format verification result of each of the sample client information
  • an association verification unit If the format verification result is passed, verify the correlation between multiple pieces of information in the sample customer information according to the correlation verification rule to obtain the correlation verification result
  • the matching verification unit is used for If the correlation verification result is passed, check whether each item of information in the sample customer information matches the preset range in the matching verification rule according to the matching verification rule to obtain a matching verification Results; a first judging unit, used for judging that the sample verification result of the sample customer information is passed if the matching verification result is passed; the second judging unit is used for if the format verification result, the If the correlation verification result or the matching verification result is not passed, it is determined that the sample verification result of the sample customer information is not passed.
  • the sample customer quantitative information acquisition unit 130 is configured to quantify the sample customer information according to a preset information quantification rule to obtain corresponding sample customer quantitative information.
  • the sample customer quantitative information acquisition unit 130 includes subunits: an item attribute information acquisition unit and an item attribute information quantification unit.
  • an item attribute information acquisition unit used for acquiring item attribute information corresponding to each of the sample customer information according to the quantification items included in the information quantification rules
  • an item attribute information quantification unit used for according to the item rules of each of the quantified items Perform quantitative processing on item attribute information corresponding to each of the sample customer information to obtain sample customer quantitative information of each sample customer information.
  • the risk verification model updating unit 140 is configured to iteratively update the preset risk verification model according to the pre-stored model update rules, the sample customer quantitative information and the sample verification result, to obtain an updated risk verification model .
  • the risk verification model update unit 140 includes subunits: an output result acquisition unit, a loss value calculation unit, a parameter update unit, a judgment unit, a determination unit, and a return execution unit.
  • the output result obtaining unit is used to obtain an output result of the quantitative information of the sample customer according to the risk verification model;
  • the loss value calculation unit is used to obtain the sample calibration of the quantitative information of the sample customer according to the loss value calculation formula.
  • the parameter update unit is used to calculate the quantitative information of the sample customer according to the gradient calculation formula, the loss value and each parameter in the risk verification model Obtain the update value of each of the parameters to update the risk verification model;
  • the judgment unit is used to determine whether each of the sample customer quantitative information has iteratively updated the risk verification model; determine
  • the unit is used to determine the risk verification model as the updated risk verification model if the quantitative information of each sample customer has been iteratively updated for the risk verification model; return to the execution unit, for if the risk verification model is updated.
  • the risk verification model is iteratively updated for each of the sample customer quantitative information, and the next sample customer quantitative information is obtained and returned to execute the described acquisition of a piece of the sample customer quantitative
  • the risk verification unit 150 is configured to perform risk verification on the newly added customer information sent by the client in real time according to the information quantification rules and the updated risk verification model, and obtain a new customer risk verification result and feedback to the client.
  • the risk verification unit 150 includes sub-units: a newly added customer quantitative information acquisition unit, a model output result acquisition unit, and a risk level matching unit.
  • a newly added customer quantitative information acquisition unit configured to quantify the newly added customer information according to the information quantification rules to obtain corresponding newly added customer quantitative information
  • a model output result acquisition unit used for acquiring according to the risk verification model The output result of the quantitative information of the newly added customer
  • a risk level matching unit configured to obtain a risk level matching the output result according to a preset risk level matching rule as the risk verification result of the newly added customer.
  • the risk level matching unit includes subunits: a verification score calculation unit and a risk level determination unit.
  • the verification score calculation unit is used to calculate and obtain the verification score according to the normalization function and the two output node values in the output result; the risk level determination unit is used to obtain the risk level interval and the verification score A level that matches the test score is used as the risk level that matches the output result.
  • the client information verification apparatus 100 further includes a subunit: a storage unit.
  • the storage unit is used for uploading the new customer risk verification result to the blockchain for storage.
  • the customer information verification device provided in the embodiment of the present application applies the above customer information verification method, randomly selects sample customer information from the historical customer information table, and performs risk verification on the sample customer information according to the risk verification rules to obtain sample verification.
  • the sample customer information is quantified to obtain the sample customer quantitative information
  • the risk verification model is iteratively updated according to the sample verification results and the sample customer quantitative information
  • the updated risk verification model is used to update the real-time data sent by the client. Perform risk verification on customer information to obtain the new customer risk verification result.
  • the administrator only needs to input the risk verification rules, and then the risk verification model can be updated in time.
  • the above-mentioned customer information verification apparatus can be implemented in the form of a computer program, and the computer program can be executed on a computer device as shown in FIG. 10 .
  • FIG. 10 is a schematic block diagram of a computer device provided by an embodiment of the present application.
  • the computer device may be a management server for executing a client information verification method to perform intelligent risk verification on client information.
  • the computer device 500 includes a processor 502 , a memory and a network interface 505 connected by a system bus 501 , wherein the memory may include a non-volatile storage medium 503 and an internal memory 504 .
  • the nonvolatile storage medium 503 can store an operating system 5031 and a computer program 5032 .
  • the computer program 5032 When executed, it can cause the processor 502 to execute the client information verification method.
  • the processor 502 is used to provide computing and control capabilities to support the operation of the entire computer device 500 .
  • the internal memory 504 provides an environment for running the computer program 5032 in the non-volatile storage medium 503.
  • the processor 502 can execute the client information verification method.
  • the network interface 505 is used for network communication, such as providing transmission of data information.
  • the network interface 505 is used for network communication, such as providing transmission of data information.
  • FIG. 10 is only a block diagram of a partial structure related to the solution of the present application, and does not constitute a limitation on the computer device 500 to which the solution of the present application is applied.
  • the specific computer device 500 may include more or fewer components than shown, or combine certain components, or have a different arrangement of components.
  • the processor 502 is configured to run the computer program 5032 stored in the memory, so as to realize the corresponding functions in the above-mentioned customer information verification method.
  • the embodiment of the computer device shown in FIG. 10 does not constitute a limitation on the specific structure of the computer device. Either some components are combined, or different component arrangements.
  • the computer device may only include a memory and a processor.
  • the structures and functions of the memory and the processor are the same as those of the embodiment shown in FIG. 10 , which will not be repeated here.
  • the processor 502 may be a central processing unit (Central Processing Unit, CPU), and the processor 502 may also be other general-purpose processors, digital signal processors (Digital Signal Processor, DSP), Application Specific Integrated Circuit (ASIC), Field-Programmable Gate Array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.
  • the general-purpose processor can be a microprocessor or the processor can also be any conventional processor or the like.
  • a computer-readable storage medium may be a non-volatile computer-readable storage medium.
  • the computer-readable storage medium stores a computer program, wherein when the computer program is executed by the processor, the steps included in the above-mentioned customer information verification method are implemented.
  • the disclosed apparatus, apparatus and method may be implemented in other manners.
  • the apparatus embodiments described above are only illustrative.
  • the division of the units is only logical function division.
  • there may be other division methods, or units with the same function may be grouped into one Units, such as multiple units or components, may be combined or may be integrated into another system, or some features may be omitted, or not implemented.
  • the shown or discussed mutual coupling or direct coupling or communication connection may be indirect coupling or communication connection through some interfaces, devices or units, and may also be electrical, mechanical or other forms of connection.
  • the units described as separate components may or may not be physically separated, and components displayed as units may or may not be physical units, that is, may be located in one place, or may be distributed to multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solutions of the embodiments of the present application.
  • each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically alone, or two or more units may be integrated into one unit.
  • the above-mentioned integrated units may be implemented in the form of hardware, or may be implemented in the form of software functional units.
  • the integrated unit if implemented in the form of a software functional unit and sold or used as an independent product, may be stored in a computer-readable storage medium.
  • the read storage medium includes several instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present application.
  • the aforementioned computer-readable storage medium includes: U disk, mobile hard disk, Read-Only Memory (ROM, Read-Only Memory), magnetic disk or optical disk and other media that can store program codes.

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Human Resources & Organizations (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Databases & Information Systems (AREA)
  • Economics (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Strategic Management (AREA)
  • General Physics & Mathematics (AREA)
  • Educational Administration (AREA)
  • Quality & Reliability (AREA)
  • Tourism & Hospitality (AREA)
  • Operations Research (AREA)
  • General Business, Economics & Management (AREA)
  • Marketing (AREA)
  • Game Theory and Decision Science (AREA)
  • Computational Linguistics (AREA)
  • Development Economics (AREA)
  • Data Mining & Analysis (AREA)
  • General Engineering & Computer Science (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

一种客户信息校验方法、装置、计算机设备及存储介质，基于人工智能技术，属于机器学习领域。所述方法包括：从历史客户信息表中随机抽取样本客户信息，根据风险校验规则对样本客户信息进行风险校验得到样本校验结果，对样本客户信息进行量化得到样本客户量化信息，根据样本校验结果及样本客户量化信息对风险校验模型进行迭代更新，并使用更新后的风险校验模型对客户端实时发出的新增客户信息进行风险校验得到新增客户风险校验结果。所述方法还涉及区块链技术，可将风险校验结果上传至区块链进行存储，管理员只需输入风险校验规则，即可及时对风险校验模型进行更新，基于更新后的风险校验模型即可对客户信息实时高效地进行风险校验。

Description

客户信息校验方法、装置、计算机设备及存储介质
本申请要求于2020年12月16日提交中国专利局、申请号为202011486655.6，发明名称为“客户信息校验方法、装置、计算机设备及存储介质”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本申请涉及人工智能技术领域，属于智慧城市中对客户信息进行智能风险校验的应用场景，尤其涉及一种客户信息校验方法、装置、计算机设备及存储介质。
背景技术
企业在进行校验业务办理过程中会获取客户信息并进行校验，以通过校验过程降低企业进行业务办理的风险，若客户信息满足相应办理条件则进行业务办理，否则表示客户信息中存在问题，企业需根据客户信息中问题的严重程度进行对应处理。传统技术方法均是采用判断方式对客户信息进行判断校验，以得到客户信息的校验结果，然而客户信息中包含关联、互斥等关系的信息，对于金融企业而言对客户信息进行判断的逻辑更是十分复杂，导致采用判断语句对客户信息进行判断耗时较长，因此发明人发现传统技术方法难以对海量并发的客户信息实时高效地进行校验，客户办理业务的过程中需等待较长时间以完成客户信息的校验，影响了企业基于客户信息进行后续业务办理的时效性。因此，现有技术方法存在无法对客户信息实时高效地进行校验的问题。
发明内容
本申请实施例提供了一种客户信息校验方法、装置、计算机设备及存储介质，旨在解决现有技术方法所存在的无法对客户信息实时高效地进行校验的问题。
第一方面，本申请实施例提供了一种客户信息校验方法，其包括：
若接收到管理员所输入的风险校验规则，从预存的历史客户信息表中随机抽取样本客户信息；
根据所述风险校验规则对所述样本客户信息进行风险校验得到每一所述样本客户信息的样本校验结果；
根据预置的信息量化规则对所述样本客户信息进行量化得到对应的样本客户量化信息；
根据预存的模型更新规则、所述样本客户量化信息及所述样本校验结果对预置的风险校验模型进行迭代更新，得到更新后的风险校验模型；
根据所述信息量化规则及所述更新后的风险校验模型对所述客户端实时发出的新增客户信息进行风险校验，得到新增客户风险校验结果并反馈至所述客户端。
第二方面，本申请实施例提供了一种客户信息校验装置，其包括：
样本客户信息获取单元，用于若接收到管理员所输入的风险校验规则，从预存的历史客户信息表中随机抽取样本客户信息；
样本校验结果获取单元，用于根据所述风险校验规则对所述样本客户信息进行风险校验得到每一所述样本客户信息的样本校验结果；
样本客户量化信息获取单元，用于根据预置的信息量化规则对所述样本客户信息进行量化得到对应的样本客户量化信息；
风险校验模型更新单元，用于根据预存的模型更新规则、所述样本客户量化信息及所述样本校验结果对预置的风险校验模型进行迭代更新，得到更新后的风险校验模型；
风险校验单元，用于根据所述信息量化规则及所述更新后的风险校验模型对所述客户端实时发出的新增客户信息进行风险校验，得到新增客户风险校验结果并反馈至所述客户端。
第三方面，本申请实施例又提供了一种计算机设备，其包括存储器、处理器及存储在所述存储器上并可在所述处理器上运行的计算机程序，所述处理器执行所述计算机程序时实现上述第一方面所述的客户信息校验方法。
第四方面，本申请实施例还提供了一种计算机可读存储介质，其中所述计算机可读存储介质存储有计算机程序，所述计算机程序当被处理器执行时使所述处理器执行上述第一方面所述的客户信息校验方法。
本申请实施例提供了一种客户信息校验方法、装置、计算机设备及存储介质。从历史客户信息表中随机抽取样本客户信息，根据风险校验规则对样本客户信息进行风险校验得到样本校验结果，对样本客户信息进行量化得到样本客户量化信息，根据样本校验结果及样本客户量化信息对风险校验模型进行迭代更新，并使用更新后的风险校验模型对客户端实时发出的新增客户信息进行风险校验得到新增客户风险校验结果。通过上述方法，管理员只需输入风险校验规则，即可及时对风险校验模型进行更新，基于更新后的风险校验模型即可对客户信息实时高效地进行风险校验。
附图说明
为了更清楚地说明本申请实施例技术方案，下面将对实施例描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图是本申请的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。
图1为本申请实施例提供的客户信息校验方法的流程示意图；
图2为本申请实施例提供的客户信息校验方法的应用场景示意图；
图3为本申请实施例提供的客户信息校验方法的子流程示意图；
图4为本申请实施例提供的客户信息校验方法的另一子流程示意图；
图5为本申请实施例提供的客户信息校验方法的另一子流程示意图；
图6为本申请实施例提供的客户信息校验方法的另一子流程示意图；
图7为本申请实施例提供的客户信息校验方法的另一子流程示意图；
图8为本申请实施例提供的客户信息校验方法的另一流程示意图；
图9为本申请实施例提供的客户信息校验装置的示意性框图；
图10为本申请实施例提供的计算机设备的示意性框图。
具体实施方式
下面将结合本申请实施例中的附图，对本申请实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例是本申请一部分实施例，而不是全部的实施例。基于本申请中的 实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本申请保护的范围。
应当理解，当在本说明书和所附权利要求书中使用时，术语“包括”和“包含”指示所描述特征、整体、步骤、操作、元素和/或组件的存在，但并不排除一个或多个其它特征、整体、步骤、操作、元素、组件和/或其集合的存在或添加。
还应当理解，在此本申请说明书中所使用的术语仅仅是出于描述特定实施例的目的而并不意在限制本申请。如在本申请说明书和所附权利要求书中所使用的那样，除非上下文清楚地指明其它情况，否则单数形式的“一”、“一个”及“该”意在包括复数形式。
还应当进一步理解，在本申请说明书和所附权利要求书中使用的术语“和/或”是指相关联列出的项中的一个或多个的任何组合以及所有可能组合，并且包括这些组合。
请参阅图1及图2，图1是本申请实施例提供的客户信息校验方法的流程示意图，图2为本申请实施例提供的客户信息校验方法的应用场景示意图，该客户信息校验方法应用于管理服务器10中，该方法通过安装于管理服务器10中的应用软件进行执行，管理服务器10与至少一台客户端20进行网络连接以实现数据信息的传输，管理服务器10即是用于执行客户信息校验方法以实现对客户信息进行智能风险校验的服务器端，管理服务器10可以是企业内所设立的服务器，管理服务器10的使用者即为企业的管理员；客户端20即是与管理服务器10建立网络连接以进行数据信息传输的终端设备，例如台式电脑、笔记本电脑、平板电脑或手机等，客户端20的使用者即为客户。如图1所示，该方法包括步骤S110～S150。
S110、若接收到管理员所输入的风险校验规则，从预存的历史客户信息表中随机抽取样本客户信息。
若接收到管理员所输入的风险校验规则，从预存的历史客户信息表中随机抽取样本客户信息。管理员可配置风险校验规则至管理服务器，其中，所输入的风险校验规则可以是全新配置得到的规则，也可以是对旧的风险校验规则进行修改后得到的规则，风险校验规则即为对客户信息进行风险校验的规则信息，风险校验规则中包含多个规则具体可采用逻辑运算符进行表示，可基于风险校验规则中的逻辑运算符对客户信息进行校验。若接收到输入的风险校验规则，则从历史客户信息表中随机抽取部分客户信息作为样本客户信息。历史客户信息表即为管理服务器中所配置的用于对历史客户信息进行存储的信息表，客户办理业务时均会通过客户端发送客户信息至管理服务器，管理服务器可将所接受到的客户信息存储至历史客户信息表中，客户信息可包含客户姓名、年龄、性别、身份证号、手机号、职业、收入、兴趣爱好、住房信息、私家车信息、住址、婚姻状态、生育信息及征信违约信息等与客户相关的信息。具体的，可根据预置的比例值从预存的历史客户信息表中随机抽取与所述比例值相匹配的样本客户信息；历史客户信息表中存储有大量客户信息，若客户信息数量较多，则可按预置的比例值从中随机抽取部分客户信息作为样本客户信息。
例如，历史客户信息表中包含一万条客户信息，预置比例值为0.1，则对应从历史客户信息表中抽取1000条客户信息作为样本客户信息。
S120、根据所述风险校验规则对所述样本客户信息进行风险校验得到每一所述样本客户 信息的样本校验结果。
根据所述风险校验规则对所述样本客户信息进行风险校验得到每一所述样本客户信息的样本校验结果。可根据风险校验规则对样本客户信息进行风险校验得到相应样本校验结果，样本校验结果可以是校验通过或校验不通过，其中，所述风险校验规则包括格式校验规则、关联校验规则及匹配校验规则。
具体的，在一实施例中，如图3所示，所述步骤S120包括子步骤：S121、S122、S123、S124和S125。
S121、根据所述格式校验规则对每一所述样本客户信息中每一项信息进行格式校验得到每一所述样本客户信息的格式校验结果。
可根据格式校验规则对样本客户信息中包含的多项信息进行格式校验，具体的，格式校验规则包含对每一项信息的字符长度、字符类型进行校验的具体规则，格式校验规则中包含与多项信息对应的校验格式，若样本客户信息中某一项信息的字符长度或字符类型不满足该项信息对应的校验格式，则该样本客户信息的格式校验结果为不通过；若样本客户信息中每一项信息的字符长度及字符类型均满足该项信息对应的校验格式，则格式校验结果为通过。
例如，身份证号这一项信息的校验格式中包括：字符长度为18个字符，前17位字符的字符类型必须是数字，最后一位字符的字符类型为数字或字母，则可根据身份证的校验格式对每一样本客户信息中与身份证相对应的一项信息进行校验。
S122、若所述格式校验结果为通过，根据所述关联校验规则对所述样本客户信息中多项信息之间的关联关系进行校验得到关联校验结果。
可根据关联校验规则对样本客户信息中多项信息之间的关联关系进行校验，具体的，关联校验规则中包括对客户信息中的多个关联信息对进行关联校验的具体规则，每一关联信息对中包含两项信息，可根据关联校验规则获取一个关联信息对所对应的两项信息的关联特征，并判断关联特征是否一致，若两项信息的关联特征一致，则表明关联信息对的两项信息存在相应关联，否则表明关联信息对的两项信息之间不存在关联。若某一样本客户信息与某一关联信息对对应的两项信息不存在关联，则该样本客户信息的关联校验结果为不通过；若某一样本客户信息与每一关联信息对对应的两项信息均存在关联，则关联校验结果为通过。
例如，某一样本客户信息中与一个关联信息对所对应的两项信息分别为“手机号：138XXXXXXXX”、“住址：A省B市C区D街道E小区”，获取到“手机号：138XXXXXXXX”的关联特征为：归属地A省B市，住址的关联特征为：所属城市A省B市，两项信息的关联特征一致，则两项信息存在关联。
S123、若所述关联校验结果为通过，根据所述匹配校验规则对所述样本客户信息中每一项信息是否与所述匹配校验规则中的预设范围相匹配进行校验得到匹配校验结果。
可根据匹配校验规则对样本客户信息中的每一项信息是否与预设范围相匹配进行校验，具体的，匹配校验规则中包含与每一项信息对应的预设范围，一项信息对应的预设范围可以是对该项信息所属范围进行限定的集合，可根据匹配校验规则判断样本客户信息中的每一项信息是否与预设范围相匹配，若某一样本客户信息中存在一项信息不与预设范围相匹配，则 该样本信息的匹配校验结果为不通过；若某一样本客户信息中每一项信息均与预设范围相匹配，则匹配校验结果为通过。
例如，某一样本客户信息中的征信违约信息为违约5次，匹配校验规则中与征信违约信息对应的预设范围为[0,3]，则该样本客户信息的征信违约次数与相应的预设范围不相匹配，该样本客户信息的匹配校验结果为不通过。
S124、若所述匹配校验结果为通过，判定所述样本客户信息的样本校验结果为通过；S125、若所述格式校验结果、所述关联校验结果或所述匹配校验结果为不通过，判定所述样本客户信息的样本校验结果为不通过。
若匹配校验结果为通过，则得到样本客户信息的样本校验结果为通过，若格式校验结果、关联校验结果及匹配校验结果中任意一个为不通过，则得到样本客户信息的样本校验结果为不通过。
S130、根据预置的信息量化规则对所述样本客户信息进行量化得到对应的样本客户量化信息。
根据预置的信息量化规则对所述样本客户信息进行量化得到对应的样本客户量化信息。信息量化规则即为对每一样本客户信息进行量化的具体规则，可根据信息量化规则对每一样本客户信息中包含的每一项信息进行量化处理，得到相应的样本客户量化信息，样本客户量化信息即可用于对每一样本客户信息进行量化表示，信息量化规则中包含多个量化项目，每一量化项目可将样本客户信息的一项信息转换为对应量化值，一个样本客户信息的多个量化值即组成为该样本客户信息的样本客户量化信息。
具体的，在一实施例中，如图4所示，步骤S130包括子步骤S131和S132。
S131、根据所述信息量化规则包含的量化项目获取每一所述样本客户信息对应的项目属性信息。
信息量化规则中可包含多个量化项目，可根据多个量化项目依次获取每一样本客户信息与每一量化项目对应的项目属性信息。
S132、根据每一所述量化项目的项目规则对每一所述样本客户信息对应的项目属性信息进行量化处理，得到每一样本客户信息的样本客户量化信息。
项目规则可对与量化项目相匹配的项目属性信息进行量化处理，具体的，每一量化项目的项目规则可将一个项目属性信息转换为一个量化值进行表示，一条样本客户量化信息可表示为一个多维的特征向量，也即是每一条项目属性信息对应样本客户量化信息中的一个维度的特征向量，对每一量化项目对应的项目属性信息进行量化所得到量化值的范围均为[0，1]。具体的，可对项目属性信息是否属于预置的特征属性进行判断，若属于特征属性则直接将该项目属性信息转换为相应特征属性值，特征属性包括身份证号、手机号等；若项目属性信息不属于特征属性，则可对项目属性信息是否为数值进行判断，若项目属性信息为数值则与该项目属性信息相匹配的项目规则为激活函数及对应的一个中间值，可通过激活函数计算得到该项目属性信息的量化值；若项目属性信息不为数值，则与该项目属性信息相匹配的项目规则包含多个关键字及与每一关键字对应的数值，获取项目规则中与项目属性信息相匹配的一 个关键字对应的数值作为该项目属性信息的量化值。
例如，若项目属性信息为特征属性，则将项目属性信息转换为对应小数进行表示。如某一样本客户信息中的身份证号为210101XXXXXXXXXXXX，则对应得到特征属性值为0.210101XXXXXXXXXXXX。
对于与量化项目对应的项目属性信息以数值方式表示的情况，对应的项目规则为一个激活函数及一个中间值，根据激活函数对中间值及该量化项目所对应的项目属性信息进行计算，即可得到对应的量化值。
例如，某一量化项目的项目规则中激活函数可表示为：
Figure PCTCN2021090587-appb-000001
其中，x为与量化项目对应的一个项目属性信息，v为项目规则包含的中间值。与收入这一量化项目对应的中间值为v＝7000，某一样本客户信息的中收入为x＝5800，则根据上述激活函数计算得到对应的量化值为0.7021。信息量化规则的婚姻状态这一量化项目对应的项目规则中包含“已婚”、“离异”、“未婚”三个关键字，与“已婚”对应的数值为“1”、与“离异”对应的数值为“0.3”，与“未婚”对应的数值为“0”，某一样本客户信息的婚姻状态为未婚，则对应的量化值为“0”。
S140、根据预存的模型更新规则、所述样本客户量化信息及所述样本校验结果对预置的风险校验模型进行迭代更新，得到更新后的风险校验模型。
根据预存的模型更新规则、所述样本客户量化信息及所述样本校验结果对预置的风险校验模型进行迭代更新，得到更新后的风险校验模型。模型更新规则即为对风险校验模型中参数值进行训练更新的规则，模型更新规则包括损失值计算公式及梯度计算公式。具体的，风险校验模型是基于神经网络所构建的智能校验模型，风险校验模型由一个输入层、多个中间层及一个输出层组成，输入层与中间层之间、中间层与前后相邻的其他中间层之间、中间层与输出层之间均通过关联公式进行关联，例如某一关联公式可表示为y＝p×x+q，p和q即为该关联公式中的参数值。输入层中包含的输入节点的数量与样本客户量化信息的维度数相等，则样本客户量化信息中的每一个量化值均与一个输入节点相对应，将多个样本客户量化信息依次输入风险校验模型进行智能校验，即可从其输出层获取输出结果，根据模型更新规则对输出结果及该样本客户量化信息对应的样本校验结果进行计算得到损失值，并根据损失值计算得到风险校验模型中每一参数值的更新值，即可对风险校验模型进行迭代更新。其中，输出结果即为输出节点的输出节点值，每一量化输入信息对应两个输出节点值，输出节点值即为样本客户量化信息与相应输出节点之间的匹配度，第一个输出节点值为校验通过的匹配度，第二个输出节点值为校验不通过的匹配度，输出节点值可采用一个小数进行表示，取值范围为[0，1]。
具体的，在一实施例中，如图5所示，步骤S140包括子步骤S141、S142、S143、S144、S145和S146。
S141、根据所述风险校验模型获取一条所述样本客户量化信息的输出结果。
根据风险校验模型获取与一条样本客户量化信息对应的输出节点值作为相应输出结果，则输出结果中包含风险校验模型两个输出节点分别对应的输出节点值。
S142、根据所述损失值计算公式获取所述样本客户量化信息的样本校验结果与所述输出结果之间的损失值。
可通过损失值计算公式对样本校验结果及输出结果进行计算，得到与该样本客户量化信息对应的损失值，损失值可用于对输出结果与样本校验结果之间的差异进行量化表示。
例如，损失函数可表示为
Figure PCTCN2021090587-appb-000002
其中，R为样本校验结果对应的结果量化值，样本校验结果为通过对应的结果量化值为“1”，样本校验结果为不通过对应的结果量化值为“0”，r 1为校验通过对应的输出节点的输出节点值，r 1为校验不通过对应的输出节点的输出节点值，f(r)即为所计算得到的损失值。
S143、根据所述梯度计算公式、所述损失值及所述风险校验模型中每一参数对所述样本客户量化信息进行计算的计算值获取每一所述参数的更新值以更新所述风险校验模型。
可根据梯度计算公式计算得到风险校验模型中每一参数的更新值以对参数原始的参数值进行更新，具体的，将风险校验模型中一个参数对一个样本客户量化信息进行计算所得到的计算值输入梯度计算公式，并结合上述损失值，即可计算得到与该参数对应的更新值，这一计算过程也即为梯度下降计算。
具体的，梯度计算公式可表示为：
Figure PCTCN2021090587-appb-000003
其中，
Figure PCTCN2021090587-appb-000004
为计算得到的参数t的更新值，ω t为参数t原始的参数值，η为梯度计算公式中预置的学习率，
Figure PCTCN2021090587-appb-000005
为基于损失值及参数t对应的计算值对该参数t的偏导值(这一计算过程中需使用参数t对应的计算值)。
S144、判断每一所述样本客户量化信息是否均已对所述风险校验模型进行迭代更新；S145、若每一所述样本客户量化信息均已对所述风险校验模型进行迭代更新，将所述风险校验模型确定为更新后的风险校验模型；S146、若每一所述样本客户量化信息未均对所述风险校验模型进行迭代更新，获取下一所述样本客户量化信息并返回执行所述根据所述风险校验模型获取一条所述样本客户量化信息的输出结果的步骤。
根据一条样本客户量化信息及对应的样本校验结果即可对风险校验模型中所有参数的参数值进行一次更新，也即是完成对风险校验模型的一次训练更新；可获取吓一条样本客户量化信息并重复上述更新过程以实现对风险校验模型进行迭代更新，直至全部样本客户量化信息均用于对风险校验模型进行迭代更新。
S150、根据所述信息量化规则及所述更新后的风险校验模型对所述客户端实时发出的新增客户信息进行风险校验，得到新增客户风险校验结果并反馈至所述客户端。
根据所述信息量化规则及所述更新后的风险校验模型对所述客户端实时发出的新增客户信息进行风险校验，得到新增客户风险校验结果并反馈至所述客户端。管理服务器可接收客户端实时发出的新增客户信息，并通过信息量化规则获取新增客户信息的新增客户量化信息，通过更新后的风险校验模型获取新增客户量化信息的输出结果，并进一步获取新增客户风险校验结果。若再次接收到管理员所输入的风险校验规则，则返回执行步骤S110中的方法。
具体的，在一实施例中，如图6所示，步骤S150包括子步骤S151、S152和S153。
S151、根据所述信息量化规则对所述新增客户信息进行量化得到对应的新增客户量化信息。
可根据信息量化规则对新增客户信息进行量化，进行量化处理的具体方式与对样本客户量化信息进行量化处理的具体方式相同，在此不作赘述。
S152、根据所述风险校验模型获取所述新增客户量化信息的输出结果。
将新增客户量化信息输入风险校验模型以获取相应输出结果，则输出结果中包含风险校验模型两个输出节点分别对应的输出节点值。
S153、根据预置的风险等级匹配规则获取与所述输出结果相匹配的风险等级作为所述新增客户风险校验结果。
根据风险等级匹配规则获取与输出结果相匹配的风险等级，得到新增客户风险校验结果。风险等级匹配规则即是用于获取与输出结果相匹配的风险等级的具体规则，风险等级越高则表示新增客户信息存在的业务风险越大。具体的，风险等级匹配规则包括归一化函数及风险等级区间。
在一实施例中，如图7所示，步骤S153包括子步骤S1531和S1532。
S1531、根据所述归一化函数及所述输出结果中两个输出节点值计算得到校验得分；S1532、获取所述风险等级区间中与所述校验得分相匹配的一个等级作为与所述输出结果相匹配的风险等级。
具体的，归一化函数可表示为：
Figure PCTCN2021090587-appb-000006
其中，r 1为校验通过对应的输出节点的输出节点值，r 1为校验不通过对应的输出节点的输出节点值，D为计算得到的校验得分。风险等级区间中包含多个等级一级每一等级对应的得分区间，获取风险等级区间中与校验得分相匹配的一个得分区间，并将该得分区间的等级作为与输出结果相匹配的风险等级。管理服务器可对新增客户风险校验结果为无风险的新增客户信息进行后续业务办理。
若新增客户风险校验结果不为无风险，则管理服务器还可在不同风险等级的新增客户信息中添加与风险等级相匹配的风险标签，并将添加风险标签的新增客户信息反馈至客户端进行修改，若新增客户风险校验结果为高风险，还可发送相应提示信息至企业员工终端以提醒企业相应员工注意风险防控。
例如，某一输出结果对应的校验得分为80.99，该校验得分与[75,90]这一得分区间相匹配，获取该得分区间的等级“低风险”作为与该输出结果相匹配的风险等级。
在一实施例中，如图8所示，步骤S150之后还包括步骤S160。
S160、将所述新增客户风险校验结果上传至区块链进行存储。
将所述新增客户风险校验结果上传至区块链进行存储，基于新增客户风险校验结果得到对应的摘要信息，具体来说，摘要信息由新增客户风险校验结果进行散列处理得到，比如利用sha256s算法处理得到。将摘要信息上传至区块链可保证其安全性和对用户的公正透明性。用户设备可以从区块链中下载得该摘要信息，以便查证新增客户风险校验结果是否被篡改。本示例所指区块链是分布式数据存储、点对点传输、共识机制、加密算法等计算机技术的新型应用模式。区块链(Blockchain)，本质上是一个去中心化的数据库，是一串使用密码学方 法相关联产生的数据块，每一个数据块中包含了一批次网络交易的信息，用于验证其信息的有效性(防伪)和生成下一个区块。区块链可以包括区块链底层平台、平台产品服务层以及应用服务层等。
在一实施例中，步骤S150之后还包括步骤：将所述新增客户信息添加至所述历史客户信息表中进行存储。
可将新增客户信息添加至历史客户信息表中，则添加至历史客户信息表中的新增客户信息可作为再次对风险校验模型进行更新的样本客户信息进行使用。
本申请中的技术方法可应用于智慧政务/智慧城管/智慧社区/智慧安防/智慧物流/智慧医疗/智慧教育/智慧环保/智慧交通等包含对客户信息进行智能风险校验的应用场景中，从而推动智慧城市的建设。
在本申请实施例所提供的客户信息校验方法中，从历史客户信息表中随机抽取样本客户信息，根据风险校验规则对样本客户信息进行风险校验得到样本校验结果，对样本客户信息进行量化得到样本客户量化信息，根据样本校验结果及样本客户量化信息对风险校验模型进行迭代更新，并使用更新后的风险校验模型对客户端实时发出的新增客户信息进行风险校验得到新增客户风险校验结果。通过上述方法，管理员只需输入风险校验规则，即可及时对风险校验模型进行更新，基于更新后的风险校验模型即可对客户信息实时高效地进行风险校验。
本申请实施例还提供一种客户信息校验装置，该客户信息校验装置用于执行前述客户信息校验方法的任一实施例。具体地，请参阅图9，图9是本申请实施例提供的客户信息校验装置的示意性框图。该客户信息校验装置可以配置于管理服务器10中。
如图9所示，客户信息校验装置100包括样本客户信息获取单元110、样本校验结果获取单元120、样本客户量化信息获取单元130、风险校验模型更新单元140和风险校验单元150。
样本客户信息获取单元110，用于样本客户信息获取单元，用于若接收到管理员所输入的风险校验规则，从预存的历史客户信息表中随机抽取样本客户信息。
样本校验结果获取单元120，用于根据所述风险校验规则对所述样本客户信息进行风险校验得到每一所述样本客户信息的样本校验结果.
在一实施例中，所述样本校验结果获取单元120包括子单元：格式校验单元、关联校验单元、匹配校验单元、第一判断单元及第二判定单元。
格式校验单元，用于根据所述格式校验规则对每一所述样本客户信息中每一项信息进行格式校验得到每一所述样本客户信息的格式校验结果；关联校验单元，用于若所述格式校验结果为通过，根据所述关联校验规则对所述样本客户信息中多项信息之间的关联关系进行校验得到关联校验结果；匹配校验单元，用于若所述关联校验结果为通过，根据所述匹配校验规则对所述样本客户信息中每一项信息是否与所述匹配校验规则中的预设范围相匹配进行校验得到匹配校验结果；第一判断单元，用于若所述匹配校验结果为通过，判定所述样本客户信息的样本校验结果为通过；第二判定单元，用于若所述格式校验结果、所述关联校验结果或所述匹配校验结果为不通过，判定所述样本客户信息的样本校验结果为不通过。
样本客户量化信息获取单元130，用于根据预置的信息量化规则对所述样本客户信息进行量化得到对应的样本客户量化信息。
在一实施例中，所述样本客户量化信息获取单元130包括子单元：项目属性信息获取单元及项目属性信息量化单元。
项目属性信息获取单元，用于根据所述信息量化规则包含的量化项目获取每一所述样本客户信息对应的项目属性信息；项目属性信息量化单元，用于根据每一所述量化项目的项目规则对每一所述样本客户信息对应的项目属性信息进行量化处理，得到每一样本客户信息的样本客户量化信息。
风险校验模型更新单元140，用于根据预存的模型更新规则、所述样本客户量化信息及所述样本校验结果对预置的风险校验模型进行迭代更新，得到更新后的风险校验模型。
在一实施例中，所述风险校验模型更新单元140包括子单元：输出结果获取单元、损失值计算单元、参数更新单元、判断单元、确定单元及返回执行单元。
输出结果获取单元，用于根据所述风险校验模型获取一条所述样本客户量化信息的输出结果；损失值计算单元，用于根据所述损失值计算公式获取所述样本客户量化信息的样本校验结果与所述输出结果之间的损失值；参数更新单元，用于根据所述梯度计算公式、所述损失值及所述风险校验模型中每一参数对所述样本客户量化信息进行计算的计算值获取每一所述参数的更新值以更新所述风险校验模型；判断单元，用于判断每一所述样本客户量化信息是否均已对所述风险校验模型进行迭代更新；确定单元，用于若每一所述样本客户量化信息均已对所述风险校验模型进行迭代更新，将所述风险校验模型确定为更新后的风险校验模型；返回执行单元，用于若每一所述样本客户量化信息未均对所述风险校验模型进行迭代更新，获取下一所述样本客户量化信息并返回执行所述根据所述风险校验模型获取一条所述样本客户量化信息的输出结果的步骤。
风险校验单元150，用于根据所述信息量化规则及所述更新后的风险校验模型对所述客户端实时发出的新增客户信息进行风险校验，得到新增客户风险校验结果并反馈至所述客户端。
在一实施例中，所述风险校验单元150包括子单元：新增客户量化信息获取单元、模型输出结果获取单元及风险等级匹配单元。
新增客户量化信息获取单元，用于根据所述信息量化规则对所述新增客户信息进行量化得到对应的新增客户量化信息；模型输出结果获取单元，用于根据所述风险校验模型获取所述新增客户量化信息的输出结果；风险等级匹配单元，用于根据预置的风险等级匹配规则获取与所述输出结果相匹配的风险等级作为所述新增客户风险校验结果。
在一实施例中，所述风险等级匹配单元包括子单元：校验得分计算单元及风险等级确定单元。
校验得分计算单元，用于根据所述归一化函数及所述输出结果中两个输出节点值计算得到校验得分；风险等级确定单元，用于获取所述风险等级区间中与所述校验得分相匹配的一个等级作为与所述输出结果相匹配的风险等级。
在一实施例中，所述客户信息校验装置100还包括子单元：存储单元。
存储单元，用于将所述新增客户风险校验结果上传至区块链进行存储。
在本申请实施例所提供的客户信息校验装置应用上述客户信息校验方法，从历史客户信息表中随机抽取样本客户信息，根据风险校验规则对样本客户信息进行风险校验得到样本校验结果，对样本客户信息进行量化得到样本客户量化信息，根据样本校验结果及样本客户量化信息对风险校验模型进行迭代更新，并使用更新后的风险校验模型对客户端实时发出的新增客户信息进行风险校验得到新增客户风险校验结果。通过上述方法，管理员只需输入风险校验规则，即可及时对风险校验模型进行更新，基于更新后的风险校验模型即可对客户信息实时高效地进行风险校验。
上述客户信息校验装置可以实现为计算机程序的形式，该计算机程序可以在如图10所示的计算机设备上运行。
请参阅图10，图10是本申请实施例提供的计算机设备的示意性框图。该计算机设备可以是用于执行客户信息校验方法以对客户信息进行智能风险校验的管理服务器。
参阅图10，该计算机设备500包括通过系统总线501连接的处理器502、存储器和网络接口505，其中，存储器可以包括非易失性存储介质503和内存储器504。
该非易失性存储介质503可存储操作系统5031和计算机程序5032。该计算机程序5032被执行时，可使得处理器502执行客户信息校验方法。
该处理器502用于提供计算和控制能力，支撑整个计算机设备500的运行。
该内存储器504为非易失性存储介质503中的计算机程序5032的运行提供环境，该计算机程序5032被处理器502执行时，可使得处理器502执行客户信息校验方法。
该网络接口505用于进行网络通信，如提供数据信息的传输等。本领域技术人员可以理解，图10中示出的结构，仅仅是与本申请方案相关的部分结构的框图，并不构成对本申请方案所应用于其上的计算机设备500的限定，具体的计算机设备500可以包括比图中所示更多或更少的部件，或者组合某些部件，或者具有不同的部件布置。
其中，所述处理器502用于运行存储在存储器中的计算机程序5032，以实现上述的客户信息校验方法中对应的功能。
本领域技术人员可以理解，图10中示出的计算机设备的实施例并不构成对计算机设备具体构成的限定，在其他实施例中，计算机设备可以包括比图示更多或更少的部件，或者组合某些部件，或者不同的部件布置。例如，在一些实施例中，计算机设备可以仅包括存储器及处理器，在这样的实施例中，存储器及处理器的结构及功能与图10所示实施例一致，在此不再赘述。
应当理解，在本申请实施例中，处理器502可以是中央处理单元(Central Processing Unit，CPU)，该处理器502还可以是其他通用处理器、数字信号处理器(Digital Signal Processor，DSP)、专用集成电路(Application Specific Integrated Circuit，ASIC)、现成可编程门阵列(Field-Programmable Gate Array，FPGA)或者其他可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬件组件等。其中，通用处理器可以是微处理器或者该处理器 也可以是任何常规的处理器等。
在本申请的另一实施例中提供计算机可读存储介质。该计算机可读存储介质可以为非易失性的计算机可读存储介质。该计算机可读存储介质存储有计算机程序，其中计算机程序被处理器执行时实现上述的客户信息校验方法中所包含的步骤。
所属领域的技术人员可以清楚地了解到，为了描述的方便和简洁，上述描述的设备、装置和单元的具体工作过程，可以参考前述方法实施例中的对应过程，在此不再赘述。本领域普通技术人员可以意识到，结合本文中所公开的实施例描述的各示例的单元及算法步骤，能够以电子硬件、计算机软件或者二者的结合来实现，为了清楚地说明硬件和软件的可互换性，在上述说明中已经按照功能一般性地描述了各示例的组成及步骤。这些功能究竟以硬件还是软件方式来执行取决于技术方案的特定应用和设计约束条件。专业技术人员可以对每个特定的应用来使用不同方法来实现所描述的功能，但是这种实现不应认为超出本申请的范围。
在本申请所提供的几个实施例中，应该理解到，所揭露的设备、装置和方法，可以通过其它的方式实现。例如，以上所描述的装置实施例仅仅是示意性的，例如，所述单元的划分，仅仅为逻辑功能划分，实际实现时可以有另外的划分方式，也可以将具有相同功能的单元集合成一个单元，例如多个单元或组件可以结合或者可以集成到另一个系统，或一些特征可以忽略，或不执行。另外，所显示或讨论的相互之间的耦合或直接耦合或通信连接可以是通过一些接口、装置或单元的间接耦合或通信连接，也可以是电的，机械的或其它的形式连接。
所述作为分离部件说明的单元可以是或者也可以不是物理上分开的，作为单元显示的部件可以是或者也可以不是物理单元，即可以位于一个地方，或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部单元来实现本申请实施例方案的目的。
另外，在本申请各个实施例中的各功能单元可以集成在一个处理单元中，也可以是各个单元单独物理存在，也可以是两个或两个以上单元集成在一个单元中。上述集成的单元既可以采用硬件的形式实现，也可以采用软件功能单元的形式实现。
所述集成的单元如果以软件功能单元的形式实现并作为独立的产品销售或使用时，可以存储在一个计算机可读存储介质中。基于这样的理解，本申请的技术方案本质上或者说对现有技术做出贡献的部分，或者该技术方案的全部或部分可以以软件产品的形式体现出来，该计算机软件产品存储在一个计算机可读存储介质中，包括若干指令用以使得一台计算机设备(可以是个人计算机，服务器，或者网络设备等)执行本申请各个实施例所述方法的全部或部分步骤。而前述的计算机可读存储介质包括：U盘、移动硬盘、只读存储器(ROM，Read-Only Memory)、磁碟或者光盘等各种可以存储程序代码的介质。
以上所述，仅为本申请的具体实施方式，但本申请的保护范围并不局限于此，任何熟悉本技术领域的技术人员在本申请揭露的技术范围内，可轻易想到各种等效的修改或替换，这些修改或替换都应涵盖在本申请的保护范围之内。因此，本申请的保护范围应以权利要求的保护范围为准。

Claims (20)

  1. 一种客户信息校验方法，应用于管理服务器中，所述管理服务器与至少一台客户端进行网络连接，其中，所述方法包括：
    若接收到管理员所输入的风险校验规则，从预存的历史客户信息表中随机抽取样本客户信息；
    根据所述风险校验规则对所述样本客户信息进行风险校验得到每一所述样本客户信息的样本校验结果；
    根据预置的信息量化规则对所述样本客户信息进行量化得到对应的样本客户量化信息；
    根据预存的模型更新规则、所述样本客户量化信息及所述样本校验结果对预置的风险校验模型进行迭代更新，得到更新后的风险校验模型；
    根据所述信息量化规则及所述更新后的风险校验模型对所述客户端实时发出的新增客户信息进行风险校验，得到新增客户风险校验结果并反馈至所述客户端。
  2. 根据权利要求1所述的客户信息校验方法，其中，所述风险校验规则包括格式校验规则、关联校验规则及匹配校验规则，所述根据所述风险校验规则对所述样本客户信息进行风险校验得到每一所述样本客户信息的样本校验结果，包括：
    根据所述格式校验规则对每一所述样本客户信息中每一项信息进行格式校验得到每一所述样本客户信息的格式校验结果；
    若所述格式校验结果为通过，根据所述关联校验规则对所述样本客户信息中多项信息之间的关联关系进行校验得到关联校验结果；
    若所述关联校验结果为通过，根据所述匹配校验规则对所述样本客户信息中每一项信息是否与所述匹配校验规则中的预设范围相匹配进行校验得到匹配校验结果；
    若所述匹配校验结果为通过，判定所述样本客户信息的样本校验结果为通过；
    若所述格式校验结果、所述关联校验结果或所述匹配校验结果为不通过，判定所述样本客户信息的样本校验结果为不通过。
  3. 根据权利要求1所述的客户信息校验方法，其中，所述根据预置的信息量化规则对所述样本客户信息进行量化得到对应的样本客户量化信息，包括：
    根据所述信息量化规则包含的量化项目获取每一所述样本客户信息对应的项目属性信息；
    根据每一所述量化项目的项目规则对每一所述样本客户信息对应的项目属性信息进行量化处理，得到每一样本客户信息的样本客户量化信息。
  4. 根据权利要求1所述的客户信息校验方法，其中，所述模型更新规则包括损失值计算公式及梯度计算公式，所述根据预存的模型更新规则、所述样本客户量化信息及所述样本校验结果对预置的风险校验模型进行迭代更新，得到更新后的风险校验模型，包括：
    根据所述风险校验模型获取一条所述样本客户量化信息的输出结果；
    根据所述损失值计算公式获取所述样本客户量化信息的样本校验结果与所述输出结果之间的损失值；
    根据所述梯度计算公式、所述损失值及所述风险校验模型中每一参数对所述样本客户量化信息进行计算的计算值获取每一所述参数的更新值以更新所述风险校验模型；
    判断每一所述样本客户量化信息是否均已对所述风险校验模型进行迭代更新；
    若每一所述样本客户量化信息均已对所述风险校验模型进行迭代更新，将所述风险校验模型确定为更新后的风险校验模型；
    若每一所述样本客户量化信息未均对所述风险校验模型进行迭代更新，获取下一所述样本客户量化信息并返回执行所述根据所述风险校验模型获取一条所述样本客户量化信息的输出结果的步骤。
  5. 根据权利要求1所述的客户信息校验方法，其中，所述根据所述信息量化规则及所述更新后的风险校验模型对所述客户端实时发出的新增客户信息进行风险校验，得到新增客户风险校验结果并反馈至所述客户端，包括：
    根据所述信息量化规则对所述新增客户信息进行量化得到对应的新增客户量化信息；
    根据所述风险校验模型获取所述新增客户量化信息的输出结果；
    根据预置的风险等级匹配规则获取与所述输出结果相匹配的风险等级作为所述新增客户风险校验结果。
  6. 根据权利要求5所述的客户信息校验方法，其中，所述风险等级匹配规则包括归一化函数及风险等级区间，所述根据预置的风险等级匹配规则获取与所述输出结果相匹配的风险等级作为所述新增客户风险校验结果，包括：
    根据所述归一化函数及所述输出结果中两个输出节点值计算得到校验得分；
    获取所述风险等级区间中与所述校验得分相匹配的一个等级作为与所述输出结果相匹配的风险等级。
  7. 根据权利要求1所述的客户信息校验方法，其中，所述根据所述信息量化规则及所述更新后的风险校验模型对所述客户端实时发出的新增客户信息进行风险校验，得到新增客户风险校验结果并反馈至所述客户端之后，还包括：
    将所述新增客户风险校验结果上传至区块链进行存储。
  8. 一种客户信息校验装置，包括：
    样本客户信息获取单元，用于若接收到管理员所输入的风险校验规则，从预存的历史客户信息表中随机抽取样本客户信息；
    样本校验结果获取单元，用于根据所述风险校验规则对所述样本客户信息进行风险校验得到每一所述样本客户信息的样本校验结果；
    样本客户量化信息获取单元，用于根据预置的信息量化规则对所述样本客户信息进行量化得到对应的样本客户量化信息；
    风险校验模型更新单元，用于根据预存的模型更新规则、所述样本客户量化信息及所述样本校验结果对预置的风险校验模型进行迭代更新，得到更新后的风险校验模型；
    风险校验单元，用于根据所述信息量化规则及所述更新后的风险校验模型对所述客户端实时发出的新增客户信息进行风险校验，得到新增客户风险校验结果并反馈至所述客户端。
  9. 一种计算机设备，包括存储器、处理器及存储在所述存储器上并可在所述处理器上运行的计算机程序，其中，所述处理器执行所述计算机程序时实现以下步骤：
    若接收到管理员所输入的风险校验规则，从预存的历史客户信息表中随机抽取样本客户信息；
    根据所述风险校验规则对所述样本客户信息进行风险校验得到每一所述样本客户信息的样本校验结果；
    根据预置的信息量化规则对所述样本客户信息进行量化得到对应的样本客户量化信息；
    根据预存的模型更新规则、所述样本客户量化信息及所述样本校验结果对预置的风险校验模型进行迭代更新，得到更新后的风险校验模型；
    根据所述信息量化规则及所述更新后的风险校验模型对所述客户端实时发出的新增客户信息进行风险校验，得到新增客户风险校验结果并反馈至所述客户端。
  10. 根据权利要求9所述的计算机设备，其中，所述风险校验规则包括格式校验规则、关联校验规则及匹配校验规则，所述根据所述风险校验规则对所述样本客户信息进行风险校验得到每一所述样本客户信息的样本校验结果，包括：
    根据所述格式校验规则对每一所述样本客户信息中每一项信息进行格式校验得到每一所述样本客户信息的格式校验结果；
    若所述格式校验结果为通过，根据所述关联校验规则对所述样本客户信息中多项信息之间的关联关系进行校验得到关联校验结果；
    若所述关联校验结果为通过，根据所述匹配校验规则对所述样本客户信息中每一项信息是否与所述匹配校验规则中的预设范围相匹配进行校验得到匹配校验结果；
    若所述匹配校验结果为通过，判定所述样本客户信息的样本校验结果为通过；
    若所述格式校验结果、所述关联校验结果或所述匹配校验结果为不通过，判定所述样本客户信息的样本校验结果为不通过。
  11. 根据权利要求9所述的计算机设备，其中，所述根据预置的信息量化规则对所述样本客户信息进行量化得到对应的样本客户量化信息，包括：
    根据所述信息量化规则包含的量化项目获取每一所述样本客户信息对应的项目属性信息；
    根据每一所述量化项目的项目规则对每一所述样本客户信息对应的项目属性信息进行量化处理，得到每一样本客户信息的样本客户量化信息。
  12. 根据权利要求9所述的计算机设备，其中，所述模型更新规则包括损失值计算公式及梯度计算公式，所述根据预存的模型更新规则、所述样本客户量化信息及所述样本校验结果对预置的风险校验模型进行迭代更新，得到更新后的风险校验模型，包括：
    根据所述风险校验模型获取一条所述样本客户量化信息的输出结果；
    根据所述损失值计算公式获取所述样本客户量化信息的样本校验结果与所述输出结果之间的损失值；
    根据所述梯度计算公式、所述损失值及所述风险校验模型中每一参数对所述样本客户量化信息进行计算的计算值获取每一所述参数的更新值以更新所述风险校验模型；
    判断每一所述样本客户量化信息是否均已对所述风险校验模型进行迭代更新；
    若每一所述样本客户量化信息均已对所述风险校验模型进行迭代更新，将所述风险校验模型确定为更新后的风险校验模型；
    若每一所述样本客户量化信息未均对所述风险校验模型进行迭代更新，获取下一所述样本客户量化信息并返回执行所述根据所述风险校验模型获取一条所述样本客户量化信息的输出结果的步骤。
  13. 根据权利要求9所述的计算机设备，其中，所述根据所述信息量化规则及所述更新后的风险校验模型对所述客户端实时发出的新增客户信息进行风险校验，得到新增客户风险校验结果并反馈至所述客户端，包括：
    根据所述信息量化规则对所述新增客户信息进行量化得到对应的新增客户量化信息；
    根据所述风险校验模型获取所述新增客户量化信息的输出结果；
    根据预置的风险等级匹配规则获取与所述输出结果相匹配的风险等级作为所述新增客户风险校验结果。
  14. 根据权利要求13所述的计算机设备，其中，所述风险等级匹配规则包括归一化函数及风险等级区间，所述根据预置的风险等级匹配规则获取与所述输出结果相匹配的风险等级作为所述新增客户风险校验结果，包括：
    根据所述归一化函数及所述输出结果中两个输出节点值计算得到校验得分；
    获取所述风险等级区间中与所述校验得分相匹配的一个等级作为与所述输出结果相匹配的风险等级。
  15. 根据权利要求9所述的计算机设备，其中，所述根据所述信息量化规则及所述更新后的风险校验模型对所述客户端实时发出的新增客户信息进行风险校验，得到新增客户风险校验结果并反馈至所述客户端之后，还包括：
    将所述新增客户风险校验结果上传至区块链进行存储。
  16. 一种计算机可读存储介质，其中，所述计算机可读存储介质存储有计算机程序，所述计算机程序当被处理器执行时使所述处理器执行以下操作：
    若接收到管理员所输入的风险校验规则，从预存的历史客户信息表中随机抽取样本客户信息；
    根据所述风险校验规则对所述样本客户信息进行风险校验得到每一所述样本客户信息的样本校验结果；
    根据预置的信息量化规则对所述样本客户信息进行量化得到对应的样本客户量化信息；
    根据预存的模型更新规则、所述样本客户量化信息及所述样本校验结果对预置的风险校验模型进行迭代更新，得到更新后的风险校验模型；
    根据所述信息量化规则及所述更新后的风险校验模型对所述客户端实时发出的新增客户信息进行风险校验，得到新增客户风险校验结果并反馈至所述客户端。
  17. 根据权利要求16所述的计算机可读存储介质，其中，所述风险校验规则包括格式校验规则、关联校验规则及匹配校验规则，所述根据所述风险校验规则对所述样本客户信息进 行风险校验得到每一所述样本客户信息的样本校验结果，包括：
    根据所述格式校验规则对每一所述样本客户信息中每一项信息进行格式校验得到每一所述样本客户信息的格式校验结果；
    若所述格式校验结果为通过，根据所述关联校验规则对所述样本客户信息中多项信息之间的关联关系进行校验得到关联校验结果；
    若所述关联校验结果为通过，根据所述匹配校验规则对所述样本客户信息中每一项信息是否与所述匹配校验规则中的预设范围相匹配进行校验得到匹配校验结果；
    若所述匹配校验结果为通过，判定所述样本客户信息的样本校验结果为通过；
    若所述格式校验结果、所述关联校验结果或所述匹配校验结果为不通过，判定所述样本客户信息的样本校验结果为不通过。
  18. 根据权利要求16所述的计算机可读存储介质，其中，所述根据预置的信息量化规则对所述样本客户信息进行量化得到对应的样本客户量化信息，包括：
    根据所述信息量化规则包含的量化项目获取每一所述样本客户信息对应的项目属性信息；
    根据每一所述量化项目的项目规则对每一所述样本客户信息对应的项目属性信息进行量化处理，得到每一样本客户信息的样本客户量化信息。
  19. 根据权利要求16所述的计算机可读存储介质，其中，所述模型更新规则包括损失值计算公式及梯度计算公式，所述根据预存的模型更新规则、所述样本客户量化信息及所述样本校验结果对预置的风险校验模型进行迭代更新，得到更新后的风险校验模型，包括：
    根据所述风险校验模型获取一条所述样本客户量化信息的输出结果；
    根据所述损失值计算公式获取所述样本客户量化信息的样本校验结果与所述输出结果之间的损失值；
    根据所述梯度计算公式、所述损失值及所述风险校验模型中每一参数对所述样本客户量化信息进行计算的计算值获取每一所述参数的更新值以更新所述风险校验模型；
    判断每一所述样本客户量化信息是否均已对所述风险校验模型进行迭代更新；
    若每一所述样本客户量化信息均已对所述风险校验模型进行迭代更新，将所述风险校验模型确定为更新后的风险校验模型；
    若每一所述样本客户量化信息未均对所述风险校验模型进行迭代更新，获取下一所述样本客户量化信息并返回执行所述根据所述风险校验模型获取一条所述样本客户量化信息的输出结果的步骤。
  20. 根据权利要求16所述的计算机可读存储介质，其中，所述根据所述信息量化规则及所述更新后的风险校验模型对所述客户端实时发出的新增客户信息进行风险校验，得到新增客户风险校验结果并反馈至所述客户端，包括：
    根据所述信息量化规则对所述新增客户信息进行量化得到对应的新增客户量化信息；
    根据所述风险校验模型获取所述新增客户量化信息的输出结果；
    根据预置的风险等级匹配规则获取与所述输出结果相匹配的风险等级作为所述新增客户风险校验结果。
PCT/CN2021/090587 2020-12-16 2021-04-28 客户信息校验方法、装置、计算机设备及存储介质 WO2022126975A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202011486655.6A CN112529429B (zh) 2020-12-16 2020-12-16 客户信息校验方法、装置、计算机设备及存储介质
CN202011486655.6 2020-12-16

Publications (1)

Publication Number Publication Date
WO2022126975A1 true WO2022126975A1 (zh) 2022-06-23

Family

ID=75000818

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/090587 WO2022126975A1 (zh) 2020-12-16 2021-04-28 客户信息校验方法、装置、计算机设备及存储介质

Country Status (2)

Country Link
CN (1) CN112529429B (zh)
WO (1) WO2022126975A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115048430A (zh) * 2022-06-28 2022-09-13 北京龙智数科科技服务有限公司 数据核验方法、系统、装置及存储介质
CN117252703A (zh) * 2023-11-20 2023-12-19 杭州联海网络科技有限公司 一种面向金融客户的营销规则生成方法和系统

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112529429B (zh) * 2020-12-16 2024-05-14 平安科技（深圳）有限公司 客户信息校验方法、装置、计算机设备及存储介质
CN113836904B (zh) * 2021-09-18 2023-11-17 唯品会(广州)软件有限公司 商品信息校验方法

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20140143134A1 (en) * 2010-01-13 2014-05-22 Corelogic Solutions, Llc System and method of detecting and assessing multiple types of risks related to mortgage lending
CN109345374A (zh) * 2018-09-17 2019-02-15 平安科技（深圳）有限公司 风险控制方法、装置、计算机设备和存储介质
CN109785117A (zh) * 2018-12-14 2019-05-21 平安科技（深圳）有限公司 基于神经网络的风控方法、计算机可读存储介质及服务器
US20200265512A1 (en) * 2019-02-20 2020-08-20 HSIP, Inc. System, method and computer program for underwriting and processing of loans using machine learning
CN111950889A (zh) * 2020-08-10 2020-11-17 中国平安人寿保险股份有限公司 一种客户风险评估方法、装置、可读存储介质及终端设备
CN112529429A (zh) * 2020-12-16 2021-03-19 平安科技（深圳）有限公司 客户信息校验方法、装置、计算机设备及存储介质

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7668776B1 (en) * 2002-01-07 2010-02-23 First Data Corporation Systems and methods for selective use of risk models to predict financial risk
CN108510396B (zh) * 2017-06-23 2020-12-29 平安科技（深圳）有限公司 投保校验的方法、装置、计算机设备及存储介质
CN107392451A (zh) * 2017-07-11 2017-11-24 重庆卡西匚匚科技有限公司 一种风险控制系统
CN107623865A (zh) * 2017-09-26 2018-01-23 武汉斗鱼网络科技有限公司 一种数据校验方法及服务器
CN109087106B (zh) * 2018-07-03 2020-12-08 创新先进技术有限公司 识别二次放号账户盗用的风控模型训练、风控方法、装置以及设备
CN109461070A (zh) * 2018-10-25 2019-03-12 深圳壹账通智能科技有限公司 一种风险审批方法、装置、存储介质和服务器
CN111861732A (zh) * 2020-07-31 2020-10-30 重庆富民银行股份有限公司 一种风险评估系统及方法

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20140143134A1 (en) * 2010-01-13 2014-05-22 Corelogic Solutions, Llc System and method of detecting and assessing multiple types of risks related to mortgage lending
CN109345374A (zh) * 2018-09-17 2019-02-15 平安科技（深圳）有限公司 风险控制方法、装置、计算机设备和存储介质
CN109785117A (zh) * 2018-12-14 2019-05-21 平安科技（深圳）有限公司 基于神经网络的风控方法、计算机可读存储介质及服务器
US20200265512A1 (en) * 2019-02-20 2020-08-20 HSIP, Inc. System, method and computer program for underwriting and processing of loans using machine learning
CN111950889A (zh) * 2020-08-10 2020-11-17 中国平安人寿保险股份有限公司 一种客户风险评估方法、装置、可读存储介质及终端设备
CN112529429A (zh) * 2020-12-16 2021-03-19 平安科技（深圳）有限公司 客户信息校验方法、装置、计算机设备及存储介质

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115048430A (zh) * 2022-06-28 2022-09-13 北京龙智数科科技服务有限公司 数据核验方法、系统、装置及存储介质
CN115048430B (zh) * 2022-06-28 2024-04-16 北京龙智数科科技服务有限公司 数据核验方法、系统、装置及存储介质
CN117252703A (zh) * 2023-11-20 2023-12-19 杭州联海网络科技有限公司 一种面向金融客户的营销规则生成方法和系统
CN117252703B (zh) * 2023-11-20 2024-02-09 杭州联海网络科技有限公司 一种面向金融客户的营销规则生成方法和系统

Also Published As

Publication number Publication date
CN112529429A (zh) 2021-03-19
CN112529429B (zh) 2024-05-14

Similar Documents

Publication Publication Date Title
WO2022126975A1 (zh) 客户信息校验方法、装置、计算机设备及存储介质
US11875400B2 (en) Systems, methods, and apparatuses for dynamically assigning nodes to a group within blockchains based on transaction type and node intelligence using distributed ledger technology (DLT)
WO2021189976A1 (zh) 一种产品信息推送方法、装置、设备及存储介质
US10133980B2 (en) Optimizing neural networks for risk assessment
US20210241120A1 (en) Systems and methods for identifying synthetic identities
US20180124071A1 (en) Collaborative content evaluation
US10346782B2 (en) Adaptive augmented decision engine
WO2019196549A1 (zh) 确定高风险用户的方法及装置
CN110866546B (zh) 一种共识节点的评估方法及装置
CN110533419B (zh) 基于区块链的风控模型的共享方法、装置及服务器
JP6060298B1 (ja) 情報配信装置、情報配信方法および情報配信プログラム
WO2022083093A1 (zh) 图谱中的概率计算方法、装置、计算机设备及存储介质
WO2022095382A1 (zh) 基于知识图谱的电子卡片生成推送方法、装置
Firmansyah et al. The influence of efficacy, credibility, and normative pressure to M-banking adoption level in Indonesia
US20230049817A1 (en) Performance-adaptive sampling strategy towards fast and accurate graph neural networks
WO2021196935A1 (zh) 数据校验方法、装置、电子设备和存储介质
CN112052891A (zh) 机器行为识别方法、装置、设备及计算机可读存储介质
KR102192949B1 (ko) 인공신경망을 이용한 창업 기업의 분석 방법 및 장치
CN114219596B (zh) 一种基于决策树模型的数据处理方法及相关设备
CN107644042B (zh) 软件程序点击率预估排序方法及服务器
US20220164374A1 (en) Method of scoring and valuing data for exchange
US20220058658A1 (en) Method of scoring and valuing data for exchange
JP2022153339A (ja) データベースシステムにおけるレコードマッチング（データベースシステムにおけるレコードマッチングのコンピュータ実装方法、コンピュータプログラム、コンピュータシステム）
CN114004456B (zh) 数据标签的计算方法、装置、计算机设备和存储介质
CN110348190B (zh) 基于用户操作行为的用户设备归属判断方法及装置

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21904914

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21904914

Country of ref document: EP

Kind code of ref document: A1