Data adjustment method and apparatus, model training method and apparatus, device, medium, and product

Abstract

The present application discloses a data adjustment method and apparatus, a model training method and apparatus, an electronic device, a medium, and a product. The data adjustment method comprises: receiving a transaction limit adjustment request sent by a user on the basis of a client; providing the client with a limit adjustment channel, and obtaining user sign information inputted on the basis of the limit adjustment channel, wherein the user sign information is acquired by means of a camera apparatus of the client; obtaining, from an identity information set, first identity information corresponding to the user sign information; if the first identity information matches second identity information in the transaction limit adjustment request, inputting the user sign information into a human body assessment model; and obtaining a human body assessment value of the user outputted by the human body assessment model, and adjusting transaction data of the user on the basis of the human body assessment value and limit adjustment data in the transaction limit adjustment request. In this way, the review efficiency of the transaction limit adjustment request is improved.

Description

Data adjustment method, model training method, device, equipment, medium and product Technical Field

[0001] This specification relates to the field of financial technology, and in particular to a data adjustment method, model training method, device, electronic equipment, medium and product. Background Art

[0002] With the development of information technology, more and more transactions can be handled online. For example, to better meet user needs, insurance companies have launched insurance coverage increase services, allowing users to request an adjustment to their insurance coverage online. However, not every customer meets the requirements for this service. Consequently, when receiving a customer's insurance coverage increase application, manual review is required to determine whether to approve the application. However, manual review is inefficient.

[0003] Summary of the Invention

[0004] The embodiments of this specification provide a data adjustment method, a model training method, an apparatus, an electronic device, a medium, and a product, which improve the efficiency of reviewing transaction limit adjustment requests.

[0005] In a first aspect, an embodiment of the present specification provides a data adjustment method, the method comprising: receiving a transaction limit adjustment request sent by a user based on a client; providing a limit adjustment channel to the client, obtaining user vital sign information input based on the limit adjustment channel, the user vital sign information being collected by a camera device based on the client; obtaining first identity information corresponding to the user vital sign information in an identity information set; if the first identity information matches the second identity information in the transaction limit adjustment request, inputting the user vital sign information into a human body assessment model; obtaining a human body assessment value of the user output by the human body assessment model, and adjusting the user's transaction data based on the human body assessment value and the limit adjustment data in the transaction limit adjustment request.

[0006] In a second aspect, an embodiment of the present specification provides an evaluation model training method, the method comprising: obtaining preset user vital signs information and a preset human body evaluation value corresponding to the preset user vital signs information; inputting the preset user vital signs information and the preset human body evaluation value into an initial training model, and training to obtain a target training model, wherein, when training the initial training model, the hyperparameter combination of the initial training model is adjusted based on the received adjustment information; obtaining verification sample data; and when the target training model is detected by using the verification sample data, determining that the target training model is a human body evaluation model, wherein the verification sample data contains multiple target user vital signs information and target human body evaluation values ​​corresponding to the target user vital signs information.

[0007] On the third aspect, an embodiment of the present specification provides a data adjustment device, which includes: a receiving unit for receiving a transaction limit adjustment request sent by a user based on a client; a first acquisition unit for providing a limit adjustment channel to the client, and acquiring user vital sign information input based on the limit adjustment channel, where the user vital sign information is collected by a camera device based on the client; a second acquisition unit for acquiring first identity information corresponding to the user vital sign information in an identity information set; an input unit for inputting the user vital sign information into a human body assessment model if the first identity information matches the second identity information in the transaction limit adjustment request; an adjustment unit for acquiring a human body assessment value of the user output by the human body assessment model, and adjusting the user's transaction data based on the human body assessment value and the limit adjustment data in the transaction limit adjustment request.

[0008] In a fourth aspect, an embodiment of the present specification provides an evaluation model training device, the device comprising: a first acquisition module for acquiring preset user vital signs information and a preset human body evaluation value corresponding to the preset user vital signs information; a training module for inputting the preset user vital signs information and the preset human body evaluation value into an initial training model to train a target training model, wherein, when training the initial training model, the adjustment parameters of the initial training model are adjusted based on different training periods; a second acquisition module for acquiring verification sample data; a determination module for determining that the target training model is a human body evaluation model when the verification sample data is used to detect that the target training model passes, wherein the verification sample data contains multiple target user vital signs information and target human body evaluation values ​​corresponding to the target user vital signs information.

[0009] In a fifth aspect, an embodiment of this specification provides a computer storage medium, which stores a plurality of instructions suitable for being loaded by a processor and executing the steps of the above method.

[0010] In a sixth aspect, an embodiment of this specification provides an electronic device, comprising: a processor and a memory; wherein the memory stores a computer program, and the computer program is suitable for being loaded by the processor and executing the steps of the above method.

[0011] In a seventh aspect, an embodiment of this specification provides a computer program product having at least one instruction stored thereon, which implements the steps of the above method when the at least one instruction is executed by a processor.

[0012] In an embodiment of the present application, after the server obtains a transaction limit adjustment request, it obtains the user's vital signs information based on the limit adjustment channel provided to the user. When it is determined through the user's vital signs information that the user's first identity information matches the second identity information in the transaction limit adjustment request, the user's vital signs information is further input into the human body evaluation model. Finally, the transaction data is adjusted based on the human body evaluation value output by the human body evaluation model and the limit adjustment data in the transaction limit adjustment request. The first identity verification determines that the user who issued the transaction limit adjustment request is the user in the transaction limit adjustment request, thereby preventing others from maliciously adjusting the transaction data and ensuring the security of modifying the transaction data. In addition, by determining whether the transaction limit adjustment request is approved based on the user's vital signs information, manual review is not required, thereby improving the review efficiency of the transaction limit adjustment request. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In order to more clearly illustrate the embodiments of this specification or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of this specification. For those skilled in the art, other drawings can be obtained based on these drawings without paying any creative work.

[0014] FIG1 is a schematic diagram of a scenario of a data adjustment method provided in an embodiment of this specification.

[0015] FIG2 is a schematic diagram of a transaction quota adjustment request provided in an embodiment of this specification.

[0016] FIG3 is a flow chart of a data adjustment method provided in an embodiment of this specification.

[0017] FIG4 is a flow chart of a data adjustment method provided in an embodiment of this specification.

[0018] FIG5 is a flow chart of a data adjustment method provided in an embodiment of this specification.

[0019] FIG6 is a schematic structural diagram of a data adjustment device provided in an embodiment of this specification.

[0020] FIG7 is a schematic diagram of the structure of an evaluation model training device provided in an embodiment of this specification.

[0021] FIG8 is a schematic structural diagram of an electronic device provided in an embodiment of this specification. DETAILED DESCRIPTION

[0022] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0023] In the description of this application, it should be understood that the terms "first", "second", etc. are used for descriptive purposes only and should not be understood to indicate or imply relative importance. In the description of this application, it should be noted that, unless otherwise expressly specified and limited, "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units that are not listed, or may optionally include other steps or units inherent to these processes, methods, products or devices. For those of ordinary skill in the art, the specific meanings of the above terms in this application can be understood according to the specific circumstances. In addition, in the description of this application, unless otherwise specified, "multiple" refers to two or more. "and / or" describes the association relationship of associated objects, indicating that three relationships can exist. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the associated objects before and after are in an "or" relationship.

[0024] Please refer to FIG1 , which is a schematic diagram of a usage scenario of the data adjustment method of this application.

[0025] In this application, a user may send a transaction limit adjustment request to a server based on a client, wherein the transaction limit adjustment request may be an insurance limit adjustment request in an embodiment of this application.

[0026] Based on this, the transaction limit adjustment request may include the transaction data identifier of the user's required limit adjustment, the second user identity information, and the limit adjustment data.

[0027] The server receives a transaction limit adjustment request, determines based on the transaction limit adjustment request that the user can adjust the limit, and then provides the client with a limit adjustment channel. Based on the limit adjustment channel, the server obtains the user's vital information and compares the user's vital information with the vital information in the identity set in the identity information set. If the comparison is successful, the first identity information corresponding to the vital information in the identity information set is determined to be the first identity information corresponding to the user's vital information. After determining that the first identity information and the second identity information match, the user's vital information is input into the human body assessment model, and the user's transaction data is adjusted based on the human body assessment value output by the human body assessment model and the limit adjustment data in the transaction limit adjustment request.

[0028] Among them, if the human body assessment value output by the human body assessment model is less than or equal to the preset assessment value, the transaction quota adjustment request initiated by the user is determined, and the user's transaction data is adjusted according to the quota adjustment data in the transaction quota adjustment request.

[0029] Refer to Figure 2, which is an example diagram of a transaction limit adjustment request for this application.

[0030] In this embodiment, the transaction limit adjustment request is initiated by the user to increase the insurance limit. Therefore, the insurance limit adjustment request in Figure 2 is the transaction limit adjustment request of this application. The insurance limit adjustment request includes the insurance policy number (equivalent to the transaction data identifier), the insured person information (equivalent to the first identity information), and the insurance limit (equivalent to the limit adjustment data).

[0031] It is understood that the information included in an insurance limit adjustment request varies depending on the type of insurance. For example, if a user submits an insurance limit adjustment request for individual insurance, both the insured and policyholder information in the insurance limit adjustment request will be primary identity information. If a user submits an insurance limit adjustment request for group insurance, the policyholder information in the insurance limit adjustment request will be primary identity information.

[0032] After obtaining the insurance limit adjustment request submitted by the user based on the client, the server provides the client with a limit adjustment channel, wherein the limit adjustment channel can be an entry for starting the client's camera device to collect the user's vital signs information, which can be a link or a button. After detecting that the user clicks the link or button based on the client, the client starts the camera device to collect the user's facial image. The user's facial image is then sent to the server. After obtaining the user's facial image, the server extracts features from the facial image to obtain the user's vital signs information, and inputs the user's vital signs information into the human body assessment model. The human body assessment model analyzes the user's current health status through facial features (user vital signs information) and outputs a health value (equivalent to the human body assessment value).

[0033] It should be noted that the information involved in the embodiments of this specification (including but not limited to user feature information) is authorized by the user or fully authorized by all parties, and the collection, use, and processing of relevant information must comply with the relevant laws, regulations, and standards of the relevant countries and regions. For example, the user feature information involved in this specification is obtained with full authorization.

[0034] It is understood that when performing feature extraction on a facial image, features such as skin color, wrinkles, acne, spots, and eye data can be extracted. The human body assessment model is a pre-trained model that can analyze the user's health value based on facial features.

[0035] Then, the server can determine, based on the health value, whether the insurance limit adjustment request submitted by the user has been approved (the health value is greater than or equal to the preset health value), and then adjust the insurance limit in the user's insurance policy according to the insurance limit in the insurance limit adjustment request, thereby completing the insurance limit adjustment. It is understandable that when using different human body assessment models, the method of determining whether to approve the transaction limit adjustment request based on the human body assessment value will be different.

[0036] Furthermore, after adjusting the insurance amount in the policy, the server can send a message of successful adjustment to the client, and send the adjusted policy to the client in electronic form for the user to view.

[0037] If the server determines that the user's health value is less than the preset health value, it sends a credit limit adjustment failure message to the client to inform the user of the adjustment result.

[0038] In an embodiment of the present application, after the server obtains a transaction limit adjustment request, it obtains the user's vital signs information based on the limit adjustment channel provided to the user. When it is determined through the user's vital signs information that the user's first identity information matches the second identity information in the transaction limit adjustment request, the user's vital signs information is further input into the human body evaluation model. Finally, the transaction data is adjusted based on the human body evaluation value output by the human body evaluation model and the limit adjustment data in the transaction limit adjustment request. The first identity verification determines that the user who issued the transaction limit adjustment request is the user in the transaction limit adjustment request, thereby preventing others from maliciously adjusting the transaction data and ensuring the security of modifying the transaction data. In addition, by determining whether the transaction limit adjustment request is approved based on the user's vital signs information, manual review is not required, thereby improving the review efficiency of the transaction limit adjustment request.

[0039] The data adjustment method provided in the embodiment of the present application will be described in detail below with reference to FIG. 3 to FIG. 8 .

[0040] Please refer to Figure 3, which is a flowchart of a data adjustment method according to an embodiment of the present application. As shown in Figure 3, the data adjustment method may include the following steps S101 to S105.

[0041] S101: Receive a transaction limit adjustment request sent by a user based on a client.

[0042] A transaction limit adjustment request may be a request to adjust the limit of a transaction, and may be a request initiated for any one of a loan limit adjustment, a credit card limit adjustment, and an insurance limit adjustment. Optionally, in the embodiment of the present application, the transaction limit adjustment request is illustrated by exemplifying an insurance limit adjustment.

[0043] A transaction limit adjustment request may include information such as the second identity information, limit adjustment data, and a transaction data identifier. The transaction data identifier may be the unique identifier of the transaction data for which the limit adjustment is required in the transaction limit adjustment request; the second identity information may be the identity information registered when applying for a transaction product, which may be a personal identification number; the limit adjustment data may be the amount of the transaction data to be adjusted; and the transaction data may be data related to the user's transaction product, such as an insurance policy.

[0044] For example, when a transaction limit adjustment request is initiated by a user to adjust an insurance limit, the transaction data identifier in the transaction limit adjustment request is the policy number of the insurance policy whose limit adjustment is currently required, and the second identity information can be the personal identification number of the insured, where the personal identification number of the insured is entered into the insurance policy corresponding to the policy number. It is understood that the data included in the transaction limit adjustment request varies for different transaction products.

[0045] Optionally, in this embodiment, the server can obtain the transaction limit adjustment request via a wireless network. The user completes the transaction limit adjustment request on the client and sends it to the server. The client can be a mobile phone, laptop, desktop computer, or other terminal.

[0046] S102: providing a credit limit adjustment channel to the client, and acquiring user vital sign information inputted based on the credit limit adjustment channel, wherein the user vital sign information is collected by a camera device based on the client.

[0047] When the server receives a transaction limit adjustment request, it provides the client with a limit adjustment channel. The limit adjustment channel can be a channel provided by the server to the user to adjust the transaction limit based on the transaction data in the transaction limit request.

[0048] After receiving the credit limit adjustment channel, the client starts the camera device to obtain the user's user vital information. The user vital information can be information describing the user's physical characteristics, which may include the user's facial features, height, weight, etc.

[0049] It should be noted that the information (including but not limited to facial features, height, weight, etc.) and data (including but not limited to personal identification numbers and credit limit adjustment data, etc.) involved in the embodiments of this specification are all authorized by the user or fully authorized by all parties, and the collection, use, and processing of relevant data must comply with the relevant laws, regulations, and standards of the relevant countries and regions. For example, the facial features, height, weight, etc. involved in this specification are all obtained with full authorization.

[0050] In the embodiment of the present application, the camera device captures the user's facial image and obtains the user's vital signs information.

[0051] In the embodiment of the present application, the facial image obtained by starting the camera device can be in the form of video data, and the user's behavior information can be obtained by analyzing continuous image frames in the video data.

[0052] S103: Obtain first identity information corresponding to the user's vital information from the identity information set.

[0053] The identity information set may include a set of multiple physical information. Similarly, the physical information may be information describing physical characteristics, which is pre-stored in the identity information set in correspondence with the first identity information. Thus, the first identity information corresponding to the user's physical information can be determined based on the identity information combined with the user's physical information.

[0054] Optionally, the identity information set can be stored on the server side. After obtaining the user's vital information, the server side directly matches the user's vital information with the vital information in the identity information set. If a match is successful, the server side determines the successfully matched vital information in the identity information set as the target vital information, obtains the identity information corresponding to the target vital information from the identity information set, and determines this identity information as the first identity information corresponding to the user's vital information. The server side directly determines the first identity information corresponding to the user's vital information using the locally stored identity information set, thereby achieving rapid determination of the first identity information.

[0055] Optionally, the identity information set may be stored in other terminals, and after receiving the user's vital information, the server sends the user's vital information to the other terminals.

[0056] After receiving the user's vital information, the other terminals compare the received user vital information with the characteristic information in the locally stored identity information set and determine the first identity information corresponding to the user vital information based on the comparison results. This determination process is similar to the server-side determination of the first identity information described above and will not be further described here. In this embodiment, by combining the identity information stored in the other terminals with the first identity information corresponding to the user vital information, user information is stored across multiple terminals, thereby improving data security.

[0057] S104: If the first identity information matches the second identity information in the transaction limit adjustment request, the user's physical sign information is input into the human body assessment model.

[0058] If the first identity information matches the second identity information in the transaction limit adjustment request, the user currently processing the transaction limit request is determined to be the same as the user whose identity information was registered when processing the transaction data, and the user's vital signs are then input into the human body assessment model. It is understood that the transaction data is the data corresponding to the transaction data identifier in the transaction limit request. For example, if the transaction limit adjustment request is for adjusting an insurance limit, the transaction data is the insurance policy number corresponding to the policy number in the transaction limit adjustment request.

[0059] If the first identity information does not match the second identity information in the transaction limit adjustment request, it is determined that there is currently a malicious modification of transaction data and an alarm is issued.

[0060] S105 , obtaining a human body evaluation value of the user output by the human body evaluation model, and adjusting the transaction data of the user based on the human body evaluation value and the quota adjustment data in the transaction quota adjustment request.

[0061] The human body assessment value is a numerical value that evaluates the current physical health status, cognitive behavioral abilities, etc. of the human body.

[0062] The human body assessment model is a pre-trained model that derives a human body assessment value based on the user's vital signs information. It can be a model for analyzing the user's health, such as an obesity assessment model (BMI model); it can also be a model for analyzing user behavior. It is understood that in the embodiments of the present application, the collected user vital signs information and the model analyzed based on the user vital signs information are adjusted based on the different transaction products corresponding to the transaction limit adjustment request.

[0063] For example, if the transaction product for a credit limit adjustment request is a credit card limit adjustment, the transaction operator needs to obtain the user's current cognitive ability. This information, including gestures, eye movements, and head movements, is input into a human body assessment model, which then determines the user's current cognitive ability.

[0064] Optionally, in an embodiment of the present application, after obtaining the human body evaluation value, the transaction data is adjusted according to the human body evaluation value and the quota adjustment data in the transaction quota adjustment request.

[0065] It can obtain the transaction data identifier in the transaction quota adjustment request, determine the preset evaluation value of the transaction data identifier, and when the human body evaluation value is less than or equal to the preset evaluation value, determine to approve the user's transaction quota adjustment request, and modify the transaction data according to the quota adjustment data in the transaction quota adjustment request. The preset evaluation value is a value that determines whether the human body evaluation value meets the preset requirements. It is understandable that the preset evaluation values ​​of different transaction products are different, but the transaction data and transaction data identifier are unique. Therefore, after obtaining the transaction data identifier, the preset evaluation value of the corresponding transaction product can be obtained, which is used as the preset evaluation value of the transaction data identifier.

[0066] For example, when the human body assessment model is a BMI model, if the user's human body assessment value (BMI value) obtained according to the output result of the BMI model is 50, and the insurance policy number in the insurance amount adjustment request is A01, its corresponding preset assessment value (preset BMI value) is 40, and the user's human body assessment value is greater than the preset assessment value, then it is determined that the user's insurance amount adjustment request will not be approved; if the user's human body assessment value is 20, which is less than the preset assessment value, then the user's insurance amount adjustment request will be approved.

[0067] Optionally, in an embodiment of the present application, in addition to setting different preset evaluation values ​​for different transaction products, different preset evaluation values ​​may also be set for different quota requests for the same transaction product.

[0068] For example, for insurance product A (a transaction product), there are three types of insurance limits: 100,000, 200,000, and 300,000. For a 100,000 limit, the corresponding preset assessment value is 40; for a 200,000 limit, the corresponding preset assessment value is 30; and for a 300,000 limit, the corresponding preset assessment value is 35. By setting different preset assessment values ​​for different insurance limits, we can accurately measure the insured's risk profile.

[0069] Optionally, in an embodiment of the present application, it can be determined whether to provide a channel to the customer based on the limit adjustment data. For example, for insurance product A, its maximum insurance limit is 300,000. When receiving a transaction limit adjustment request sent by the user based on the client, it is determined that the limit adjustment data in the transaction limit adjustment request is greater than 300,000, and a limit adjustment channel is provided to the client. Afterwards, after receiving the user's vital signs information sent by the user based on the limit adjustment channel, the user's human body evaluation value obtained based on the user's vital signs information is less than the preset evaluation value corresponding to the maximum insurance limit of the insurance product A. When it is determined that the user's insurance limit adjustment request is approved, the limit adjustment data in the user's transaction limit adjustment request is changed to the maximum insurance limit, that is, changed to 300,000. The user's insurance policy is updated based on the limit adjustment data.

[0070] In an embodiment of the present application, whether the user's transaction limit adjustment request is approved is determined by the preset evaluation value of the transaction product and the human body evaluation value output by the human body evaluation model. The setting of the preset evaluation value is related to the transaction product, which improves the accuracy of judging whether the user's transaction limit adjustment request is approved.

[0071] In an embodiment of the present application, after the server obtains a transaction limit adjustment request, it obtains the user's vital information based on the limit adjustment channel provided to the user. When it is determined through the user's vital information that the user's first identity information matches the second identity information in the transaction limit adjustment request, the user's vital information is further input into a human body assessment model, and then the transaction data is adjusted based on the human body assessment value output by the human body assessment model and the limit adjustment data in the transaction limit adjustment request. The first identity verification determines that the user who issued the transaction limit adjustment request is the user in the transaction limit adjustment request, thereby preventing others from maliciously adjusting the transaction data and ensuring the security of modifying the transaction data. In addition, by determining whether the transaction limit adjustment request is approved based on the user's vital information, manual review is not required, thereby improving the review efficiency of the transaction limit adjustment request.

[0072] Referring to Figure 4, which is a flow chart of an embodiment of the present application, the data adjustment method includes steps S201 to S206.

[0073] S201: Obtain second identity information in a transaction limit adjustment request.

[0074] S202: Obtain the transaction data identifier in the transaction limit adjustment request, and determine the risk factor of the transaction product corresponding to the transaction data identifier.

[0075] S203: Determine the risk level of the user based on the risk factors and the second identity information.

[0076] S204: When the risk level is less than or equal to a preset level, determine the maximum amount of the transaction product in different dimensions, wherein there is at least one maximum amount.

[0077] S205: Determine the lowest limit among the highest limits as the target limit, and obtain the limit adjustment data in the transaction limit adjustment request.

[0078] S206: When the credit limit adjustment data is less than the target credit limit, a credit limit adjustment channel is provided to the client, and user vital sign information input based on the credit limit adjustment channel is obtained.

[0079] In the embodiment of the present application, for each transaction of the transaction product, there is corresponding transaction data, which is the certificate for handling the transaction, including the transaction data identifier, the second identity information, etc.

[0080] The determination information may be information collected to determine the user's risk level. The risk factors may be the risks associated with each type of transaction product. Different transaction products may have different risk factors.

[0081] For example, when the transaction product is an insurance-type transaction product, for health insurance, it is usually necessary to obtain the insured's health status information as a risk factor, which may include height, weight, family medical history, past illnesses, physical examination records, and historical claims records: none. For car insurance, it is usually necessary to obtain the owner and vehicle information as risk factors, including vehicle brand, model, year, usage, and driver information. For life insurance, it is usually necessary to obtain the insured's personal information and medical records as risk factors, including age, physical condition, family medical history, and smoking status.

[0082] It should be noted that the information involved in the embodiments of this specification (including but not limited to age, marital status, occupation, etc.) is authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with the relevant laws, regulations and standards of relevant countries and regions.

[0083] It is understandable that in the embodiment of the present application, there are multiple determination information collected based on the second identity information. For example, it can be the user's age: 42 years old, height: 165 cm, weight 50 kg, family medical history: none, past illness: none, medical record: healthy, occupation: teacher, historical claims record: none. The transaction product determined by the transaction data identifier of the transaction limit adjustment request is health insurance, and the risk factors for determining the product include height, weight, family medical history, physical examination records, and historical claims records. Then, the target determination information is determined in the determination information based on the risk factors. Optionally, the determination information determined in this embodiment includes: age: 42 years old, height: 165 cm, weight 50 kg, family medical history: none, past illness: none, physical examination records: healthy, and historical claims record: none. The user's risk level is then determined based on the target determination information. Among them, the determination of the user's risk level based on the target determination information can be determined by weight.

[0084] The preset level can be used to determine whether a user's risk level meets the pre-set criteria for adjusting transaction limits. If the user's risk level is determined to be less than or equal to the preset level, the maximum limit for the transaction product under different dimensions is further determined.

[0085] The dimensions include the platform dimension and the transaction product company dimension. The platform dimension refers to the adjustment limit that the platform can provide, while the transaction product company dimension refers to the adjustment limit that the transaction product company can provide. Generally speaking, the maximum limit of the platform is greater than that of the insurance company. This is because the platform, as an intermediary, has the primary goal of attracting more customers to promote transaction development.

[0086] In this embodiment, when it is determined that the risk level of the user is less than or equal to the preset level, it is necessary to further review the transaction limit adjustment request initiated by the user to determine whether to provide the user with a limit adjustment channel.

[0087] The determination method can be to obtain the transaction data identifier in the transaction limit adjustment request, and determine the maximum limit of the transaction product corresponding to the transaction data identifier in different dimensions. It can be understood that the transaction product has a maximum limit in each dimension, and thus, the transaction product has multiple maximum limits in different dimensions. The lowest limit is determined as the target limit among the maximum limits. The lowest limit is compared with the limit adjustment data that the user wants to adjust. When the limit adjustment data is less than the target limit, it is determined that a limit adjustment channel can be provided to the user to adjust the user's transaction data. In this embodiment, based on the maximum limit of the transaction product in different dimensions and the limit adjustment data submitted by the user, a basis is provided for determining whether to provide a limit adjustment channel to the user.

[0088] In this embodiment, the transaction product of the transaction data that the user needs to adjust is determined according to the transaction limit adjustment request, and the corresponding target judgment factor is obtained based on the transaction product to determine the user's risk level, thereby achieving accurate determination of the user's risk level.

[0089] In this embodiment, when it is determined that the user's risk level is lower than the preset level, it is further determined whether to provide the user with the credit limit adjustment data based on the maximum credit limit of the transaction product in different dimensions and the credit limit adjustment data in the transaction credit limit adjustment request submitted by the user, providing a basis for determining whether to provide the user with a credit limit adjustment channel.

[0090] 5 , which is a schematic diagram of an implementation flow of the evaluation model training method of the present application, includes steps S1 to S4 .

[0091] S1, obtaining preset user physical sign information and a preset human body assessment value corresponding to the preset user physical sign information.

[0092] The preset user physical sign information may be the user physical sign information in the sample data for training the initial training model, which may include the user's facial feature information, height, weight and other information.

[0093] The preset human body evaluation value may be a numerical value used to evaluate the user's physical health status, cognitive behavioral ability, etc. in the sample data used to train the initial training model.

[0094] S2, inputting the preset user physical sign information and the preset human body assessment value into the initial training model, and training to obtain the target training model, wherein, when training the initial training model, the hyperparameter combination of the initial training model is determined based on the received adjustment information.

[0095] The initial training model may be a model that can be trained to determine the human body evaluation value. The target training model may be a model that can be used to determine the human body evaluation value.

[0096] In the present application, preset user vital sign information and preset human body assessment values ​​associated with the preset user vital sign information are input into an initial training model, and the initial training model is trained to obtain a target training model.

[0097] Hyperparameter combinations can be parameters that need to be adjusted during the initial model training process to optimize model performance. For example, they can be optimizer parameters, such as learning rate, momentum, weight decay, etc.

[0098] The adjustment information is information obtained based on user input and used to adjust the performance of the initial training model.

[0099] When training the initial training model, you can first define a set of value ranges for hyperparameter combinations, and then input preset user vital sign information and preset human body assessment values ​​into the initial training model to train and obtain the target training model.

[0100] Optionally, in this embodiment, the preset user vital signs information can be first detected using a detection model. When it is determined that the preset user vital signs information passes the detection based on the output result of the detection model, the preset user vital signs information and the preset human body evaluation value are input into the initial training model.

[0101] The detection model may be a model for detecting whether the captured preset user vital signs information meets the conditions, for example, detecting whether the preset user vital signs information is comprehensive.

[0102] In this embodiment, the quality of the preset user vital signs information is judged by the detection model, and based on the judgment result, it is determined whether to input the preset user vital signs information and its corresponding preset human body evaluation value into the initial training model, thereby improving the accuracy of the human body evaluation model trained according to the initial training model.

[0103] S3, obtain verification sample data.

[0104] Validation sample data is data used to verify whether the target training model is accurate.

[0105] S4, when the target training model passes the detection using the verification sample data, determining that the target training model is a human body evaluation model, wherein the verification sample data includes multiple target user physical sign information and target human body evaluation values ​​corresponding to the target user physical sign information.

[0106] The target user's physical information may be the user's physical information in the verification sample data, which may include the user's facial feature information, height, weight, and other information.

[0107] The target human body evaluation value can be a numerical value in the verification sample data that evaluates the user's physical health status, cognitive behavioral ability, etc.

[0108] In this embodiment, the target user's vital signs information can be input into the target training model, and the output result of the target training model can be compared with the target human body evaluation value corresponding to the target user's vital signs information. If the comparison result is within a preset range, the target training model is determined to be a human body evaluation model, wherein the preset range can be a basis for error judgment, and its specific value can be determined by experience.

[0109] If it is determined based on the verification sample data that the target training model has not passed, the hyperparameter combination of the target training model is changed according to the adjustment information input by the user until it is detected that the target training model has passed, thereby obtaining a human body evaluation model.

[0110] In this example, when training the human body evaluation model, the accuracy of the trained human body evaluation model is improved by adjusting the training model according to the adjustment parameters.

[0111] Please refer to Figure 6, which provides a structural diagram of a data adjustment device according to an embodiment of the present specification. As shown in Figure 6, the data adjustment device 1 according to the embodiment of the present specification may include: a receiving unit 11, a first obtaining unit 12, a second obtaining unit 13, an input unit 14, and an adjustment unit 15. The receiving unit 11 is used to receive a transaction limit adjustment request sent by a user based on a client; the first obtaining unit 12 is used to provide a limit adjustment channel to the client, and obtain the user's vital signs information input based on the limit adjustment channel, and the user's vital signs information is collected by a camera device based on the client; the second obtaining unit 13 is used to obtain the first identity information corresponding to the user's vital signs information in the identity information set; the input unit 14 is used to input the user's vital signs information into the human body assessment model if the first identity information matches the second identity information in the transaction limit adjustment request; the adjustment unit 15 is used to obtain the user's human body assessment value output by the human body assessment model, and adjust the user's transaction data based on the human body assessment value and the limit adjustment data in the transaction limit adjustment request.

[0112] In a possible embodiment, the first acquisition unit 12 is used to obtain the second identity information in the transaction limit adjustment request; obtain the transaction data identifier in the transaction limit adjustment request, and determine the risk factor of the transaction product corresponding to the transaction data identifier; determine the risk level of the user based on the risk factor and the second identity information; when the risk level is less than or equal to the preset level, determine the maximum limit of the transaction product in different dimensions, wherein there is at least one maximum limit; determine the lowest limit among the maximum limits as the target limit, and obtain the limit adjustment data in the transaction limit adjustment request; when the limit adjustment data is less than the target limit, provide a limit adjustment channel to the client, and obtain user vital sign information input based on the limit adjustment channel.

[0113] In a possible implementation, the first acquisition unit 12 is configured to acquire determination information of the user based on the second identity information; determine target determination information corresponding to the risk factor in the determination information; and determine the risk level of the user based on the target determination information.

[0114] In one possible implementation, the adjustment unit 15 is used to determine a preset evaluation value of the transaction product based on the transaction data identifier in the transaction quota adjustment request; when the human body evaluation value is less than or equal to the preset evaluation value, determine to approve the user's transaction quota adjustment request; and adjust the user's transaction data based on the request data in the transaction quota adjustment.

[0115] In one possible embodiment, the second acquisition unit 13 is used to compare the user's vital information with the vital information in the identity information set; if there is target vital information in the identity information set that is consistent with the user's vital information, the identity information corresponding to the target vital information is determined as the first identity information corresponding to the user's vital information.

[0116] In a possible implementation, the second acquiring unit 13 is configured to send the user vital sign information to other terminals; and receive first identity information sent by other terminals, wherein the first identity information is determined based on a set of identity information stored in other terminals.

[0117] In an embodiment of the present application, after the server obtains a transaction limit adjustment request, it obtains the user's vital signs information based on the limit adjustment channel provided to the user. When it is determined through the user's vital signs information that the user's first identity information matches the second identity information in the transaction limit adjustment request, the user's vital signs information is further input into the human body evaluation model. Finally, the transaction data is adjusted based on the human body evaluation value output by the human body evaluation model and the limit adjustment data in the transaction limit adjustment request. The first identity verification determines that the user who issued the transaction limit adjustment request is the user in the transaction limit adjustment request, thereby preventing others from maliciously adjusting the transaction data and ensuring the security of modifying the transaction data. In addition, by determining whether the transaction limit adjustment request is approved based on the user's vital signs information, manual review is not required, thereby improving the review efficiency of the transaction limit adjustment request.

[0118] Please refer to Figure 7, which provides a structural diagram of an evaluation model training device according to an embodiment of the present specification. As shown in Figure 7, the evaluation model training device 2 according to the embodiment of the present specification may include: a first acquisition module 21 for acquiring preset user vital signs information and preset human body evaluation values ​​corresponding to the preset user vital signs information; a training module 22 for inputting the preset user vital signs information and the preset human body evaluation values ​​into the initial training model to train a target training model, wherein, when training the initial training model, the hyperparameter combination of the initial training model is adjusted based on the received adjustment information; a second acquisition module 23 for acquiring verification sample data; a determination module 24 for determining that the target training model is a human body evaluation model when the verification sample data is used to detect that the target training model passes, wherein the verification sample data includes multiple target user vital signs information and target human body evaluation values ​​corresponding to the target user vital signs information.

[0119] In one possible implementation, the training module 22 is used to detect preset user vital signs information using a detection model; when it is determined that the preset user vital signs information passes the detection based on the output result of the detection model, the preset user vital signs information and the preset human body evaluation value are input into the initial training model to train and obtain a target training model.

[0120] In this example, when training the human body evaluation model, the adjustment parameters of the initial training model are dynamically adjusted in different training periods so that the adjustment parameters of the initial training model correspond to the training period, thereby improving the accuracy of the trained human body evaluation model.

[0121] The embodiments of this specification also provide a computer storage medium, which can store multiple program instructions. The program instructions are suitable for being loaded by a processor and executing the method steps of the embodiments shown in Figures 1 to 5 above. The specific execution process can refer to the specific description of the embodiments shown in Figures 1 to 5, which will not be repeated here.

[0122] Please refer to Figure 8, which provides a structural diagram of an electronic device for an embodiment of this specification. As shown in Figure 8, the electronic device 1000 may include: at least one processor 1001, such as a CPU, at least one network interface 1004, a user interface 1003, a memory 1005, and at least one communication bus 1002. Among them, the communication bus 1002 is used to realize the connection and communication between these components. Among them, the network interface 1004 may optionally include a standard wired interface, a wireless interface (such as a WI-FI interface). The memory 1005 may be a high-speed RAM memory, or a non-volatile memory (non-volatile memory), such as at least one disk storage. The memory 1005 may also be at least one storage device located away from the aforementioned processor 1001. As shown in Figure 8, the memory 1005 as a computer storage medium may include an operating system, a network communication module, an input and output interface module, and a data adjustment program.

[0123] In the electronic device 1000 shown in FIG8 , the user interface 1003 is mainly used to provide an input interface for the user and obtain data input by the user.

[0124] In one embodiment, the processor 1001 can be used to call the data adjustment program stored in the memory 1005, and specifically perform the following operations: receive a transaction limit adjustment request sent by a user based on a client; provide a limit adjustment channel to the client, obtain user vital sign information input based on the limit adjustment channel, and the user vital sign information is collected by a camera device based on the client; obtain the first identity information corresponding to the user vital sign information in the identity information set; if the first identity information matches the second identity information in the transaction limit adjustment request, input the user vital sign information into the human body assessment model; obtain the user's human body assessment value output by the human body assessment model, and adjust the user's transaction data based on the human body assessment value and the limit adjustment data in the transaction limit adjustment request.

[0125] In one embodiment, when the processor 1001 provides a credit limit adjustment channel to the client and obtains user vital sign information input based on the credit limit adjustment channel, it also performs the following operations: obtains the second identity information in the transaction credit limit adjustment request; obtains the transaction data identifier in the transaction credit limit adjustment request, and determines the risk factor of the transaction product corresponding to the transaction data identifier; determines the risk level of the user based on the risk factor and the second identity information; when the risk level is less than or equal to the preset level, determines the maximum credit limit of the transaction product in different dimensions, wherein there is at least one maximum credit limit; determines the lowest credit limit among the maximum credit limits as the target credit limit, and obtains the credit limit adjustment data in the transaction credit limit adjustment request; when the credit limit adjustment data is less than the target credit limit, provides the credit limit adjustment channel to the client, and obtains the user vital sign information input based on the credit limit adjustment channel.

[0126] In one embodiment, the processor 1001 determines the risk level of the user based on the risk factors and the second identity information, and further performs the following operations: obtaining the judgment information of the user based on the second identity information; determining the target judgment information corresponding to the risk factors in the judgment information; and determining the risk level of the user based on the target judgment information.

[0127] In one embodiment, the processor 1001 adjusts the user's transaction data based on the human body evaluation value and the quota adjustment data in the transaction quota adjustment request, and also performs the following operations: determining the preset evaluation value of the transaction product according to the transaction data identifier in the transaction quota adjustment request; when the human body evaluation value is less than or equal to the preset evaluation value, determining to approve the user's transaction quota adjustment request; and adjusting the user's transaction data according to the request data in the transaction quota adjustment.

[0128] In one embodiment, when the processor 1001 obtains the first identity information corresponding to the user vital signs information in the identity information set, it also performs the following operations: compares the user vital signs information with the vital signs information in the identity information set; if there is target vital signs information consistent with the user vital signs information in the identity information set, the identity information corresponding to the target vital signs information is determined to be the first identity information corresponding to the user vital signs information.

[0129] In one embodiment, when the processor 1001 obtains the first identity information corresponding to the user vital information in the identity information set, it also performs the following operations: sending the user vital information to other terminals; receiving the first identity information sent by other terminals, wherein the first identity information is determined based on the identity information set stored in the other terminals.

[0130] In one embodiment, the processor 1001 further performs the following operations: obtaining preset user vital signs information and preset human body evaluation values ​​corresponding to the preset user vital signs information; inputting the preset user vital signs information and the preset human body evaluation values ​​into the initial training model, and training to obtain a target training model, wherein, when training the initial training model, the adjustment parameters of the initial training model are adjusted based on different training periods; obtaining verification sample data; and determining that the target training model is a human body evaluation model when the verification sample data is used to detect that the target training model passes, wherein the verification sample data contains multiple target user vital signs information and target human body evaluation values ​​corresponding to the target user vital signs information.

[0131] In one embodiment, the processor 1001 inputs the preset user vital signs information and the preset human body evaluation value into the initial training model to train a target training model, and further performs the following operations: uses the detection model to detect the preset user vital signs information; when it is determined that the preset user vital signs information has passed the detection based on the output result of the detection model, the preset user vital signs information and the preset human body evaluation value are input into the initial training model to train a target training model.

[0132] In an embodiment of the present application, after the server obtains a transaction limit adjustment request, it obtains the user's vital signs information based on the limit adjustment channel provided to the user. When it is determined through the user's vital signs information that the user's first identity information matches the second identity information in the transaction limit adjustment request, the user's vital signs information is further input into the human body evaluation model. Finally, the transaction data is adjusted based on the human body evaluation value output by the human body evaluation model and the limit adjustment data in the transaction limit adjustment request. The first identity verification determines that the user who issued the transaction limit adjustment request is the user in the transaction limit adjustment request, thereby preventing others from maliciously adjusting the transaction data and ensuring the security of modifying the transaction data. In addition, by determining whether the transaction limit adjustment request is approved based on the user's vital signs information, manual review is not required, thereby improving the review efficiency of the transaction limit adjustment request.

[0133] The embodiments of this specification provide a computer program product having at least one instruction stored thereon, which implements the steps of the above embodiments when the at least one instruction is executed by a processor.

[0134] Those skilled in the art will appreciate that all or part of the processes in the above-described method embodiments can be implemented by instructing related hardware through a computer program. The program can be stored in a computer-readable storage medium, and when executed, the program can include the processes in the above-described method embodiments. The storage medium can be a magnetic disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

[0135] The above disclosure is only a preferred embodiment of this specification, and certainly cannot be used to limit the scope of rights of this specification. Therefore, equivalent changes made according to the claims of this specification are still within the scope covered by this specification.

Claims

1. A data adjustment method, applied to a server, comprising: Receive transaction quota adjustment requests sent by users based on the client; Providing a credit limit adjustment channel to the client, and obtaining user vital sign information input based on the credit limit adjustment channel, wherein the user vital sign information is collected based on a camera device of the client; Acquire first identity information corresponding to the user's vital sign information from the identity information set; If the first identity information matches the second identity information in the transaction limit adjustment request, inputting the user's physical sign information into a human body assessment model; A human body evaluation value of the user output by the human body evaluation model is obtained, and transaction data of the user is adjusted based on the human body evaluation value and the quota adjustment data in the transaction quota adjustment request.

2. According to the method of claim 1, providing a credit limit adjustment channel to the client and obtaining user vital sign information input based on the credit limit adjustment channel comprises: Obtaining the second identity information in the transaction limit adjustment request; Obtaining the transaction data identifier in the transaction quota adjustment request, and determining the risk factor of the transaction product corresponding to the transaction data identifier; determining a risk level of the user based on the risk factor and the second identity information; When the risk level is less than or equal to the preset level, determining the maximum amount of the transaction product in different dimensions; Determining the lowest amount among the highest amounts as the target amount, and obtaining the amount adjustment data in the transaction amount adjustment request; When the credit limit adjustment data is less than the target credit limit, a credit limit adjustment channel is provided to the client, and user vital sign information input based on the credit limit adjustment channel is obtained.

3. The method according to claim 2, wherein determining the risk level of the user based on the risk factor and the second identity information comprises: acquiring determination information of the user according to the second identity information; determining target determination information corresponding to the risk factor in the determination information; The risk level of the user is determined according to the target determination information.

4. The method according to claim 1, wherein adjusting the transaction data of the user based on the human body evaluation value and the quota adjustment data in the transaction quota adjustment request comprises: Determining a preset valuation value of a transaction product according to the transaction data identifier in the transaction quota adjustment request; When the human body evaluation value is less than or equal to the preset evaluation value, it is determined that the transaction amount adjustment of the user Complete request; The transaction data of the user is adjusted according to the request data in the transaction quota adjustment.

5. The method according to claim 1, wherein the identity information set is stored on the server side, and the step of obtaining the first identity information corresponding to the user's vital sign information from the identity information set comprises: Comparing the user's vital sign information with the vital sign information in the identity information set; If there is target vital sign information consistent with the user vital sign information in the identity information set, the identity information corresponding to the target vital sign information is determined to be the first identity information corresponding to the user vital sign information.

6. The method according to claim 1, wherein the identity information set is stored in other terminals, and the step of obtaining the first identity information corresponding to the user vital sign information from the identity information set comprises: Sending the user's vital sign information to the other terminal; Receive first identity information sent by the other terminal, wherein the first identity information is determined based on a set of identity information stored in the other terminal.

7. A method for training an evaluation model, the method comprising: Obtaining preset user vital sign information and a preset human body assessment value corresponding to the preset user vital sign information; Inputting the preset user vital sign information and the preset human body assessment value into an initial training model, and training to obtain a target training model, wherein, when training the initial training model, a hyperparameter combination of the initial training model is adjusted based on the received adjustment information; Obtain verification sample data; When the target training model is detected by using the verification sample data, it is determined that the target training model is a human body assessment model, wherein the verification sample data includes multiple target user vital signs information and target human body assessment values ​​corresponding to the target user vital signs information.

8. The method according to claim 7, wherein the inputting the preset user vital sign information and the preset human body evaluation value into the initial training model to train to obtain a target training model comprises: Using a detection model to detect the preset user vital sign information; When it is determined according to the output result of the detection model that the preset user vital sign information passes the detection, the preset user vital sign information and the preset human body evaluation value are input into the initial training model to train and obtain a target training model.

9. A data adjustment device, comprising: A receiving unit, configured to receive a transaction quota adjustment request sent by a user based on a client; A first acquisition unit is used to provide a credit limit adjustment channel to the client, and acquire user vital sign information input based on the credit limit adjustment channel, wherein the user vital sign information is collected based on a camera device of the client; A second acquisition unit, used to acquire first identity information corresponding to the user's vital sign information from the identity information set; An input unit, configured to input the user's physical sign information into a human body assessment model if the first identity information matches the second identity information in the transaction limit adjustment request; An adjustment unit is used to obtain a human body evaluation value of the user output by the human body evaluation model, and adjust the transaction data of the user based on the human body evaluation value and the quota adjustment data in the transaction quota adjustment request.

10. An evaluation model training device, comprising: A first acquisition module, used to acquire preset user vital sign information and a preset human body assessment value corresponding to the preset user vital sign information; A training module, used for inputting the preset user vital sign information and the preset human body evaluation value into an initial training model, and training to obtain a target training model, wherein when training the initial training model, the adjustment parameters of the initial training model are adjusted based on different training periods; The second acquisition module is used to acquire verification sample data; A determination module is used to determine that the target training model is a human body assessment model when the target training model is detected by using the verification sample data, wherein the verification sample data includes multiple target user vital signs information and target human body assessment values ​​corresponding to the target user vital signs information.

11. A computer storage medium storing a plurality of instructions, wherein the instructions are suitable for being loaded by a processor and executing the steps of the method according to any one of claims 1 to 8.

12. An electronic device having at least one instruction stored thereon, wherein when the at least one instruction is executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.

13. A computer program product having at least one instruction stored thereon, wherein when the at least one instruction is executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.

Images