CN113837595A

CN113837595A - Surface label mode distribution method, device, equipment and storage medium

Info

Publication number: CN113837595A
Application number: CN202111106073.5A
Authority: CN
Inventors: 李欣; 徐端琦; 李鹏; 胡君一; 王大森; 罗佳; 韩韬; 杨绪森
Original assignee: China Unionpay Co Ltd
Current assignee: China Unionpay Co Ltd
Priority date: 2021-09-22
Filing date: 2021-09-22
Publication date: 2021-12-24

Abstract

The application discloses a method, a device, equipment and a storage medium for allocating a face-to-face label mode, and belongs to the field of data processing. The method comprises the following steps: responding to the received surface label application message, and obtaining model input data according to the obtained user basic data and service related data of the surface label application user; inputting model input data into a pre-trained surface label decision model to obtain an output result corresponding to a surface label application user, wherein the output result is used for representing a surface label mode applicable to the surface label application user, the surface label mode comprises an on-line label mode or an off-line label mode, the surface label decision model is obtained by utilizing a multi-dimensional classification algorithm for training according to first sample data of a plurality of sample users, and the first sample data comprises sample user basic data, sample service related data and an expected surface label mode; and according to the output result, allocating a face label mode corresponding to the output result for the face label application user. According to the embodiment of the application, the efficiency of the surface label application can be improved.

Description

Surface label mode distribution method, device, equipment and storage medium

Technical Field

The present application relates to the field of data processing, and in particular, to a method, an apparatus, a device, and a storage medium for allocating a face-to-face label.

Background

With the development of internet technology and financial technology, it is difficult for traditional offline business transaction to meet the service requirements of customers, so online business transaction becomes a great development trend.

Some financial business needs surface labels, which can be classified into an on-line label and an off-line label. The online surface label can be a remote surface label realized by using an audio and video communication technology. The under-line facing sign is a facing sign that requires the user to go to the financial institution personally. At the present stage, the user needs to select the on-line label or the under-line label through manual customer guidance or page guidance selection and the like, so that the time spent by the user for applying the surface label is long, and the efficiency of applying the surface label is low.

Disclosure of Invention

The embodiment of the application provides a method, a device, equipment and a storage medium for allocating surface label modes, and can improve the efficiency of surface label application.

In a first aspect, an embodiment of the present application provides a face-to-face label allocation method, including: responding to the received surface label application message, and obtaining model input data according to the obtained user basic data and service related data of the surface label application user, wherein the surface label application message is sent by the service provider equipment in response to the surface label application of the user terminal of the surface label application user; inputting model input data into a pre-trained surface label decision model to obtain an output result corresponding to a surface label application user, wherein the output result is used for representing a surface label mode applicable to the surface label application user, the surface label mode comprises an on-line label mode or an off-line label mode, the surface label decision model is obtained by utilizing a multi-dimensional classification algorithm for training according to first sample data of a plurality of sample users, and the first sample data comprises sample user basic data, sample service related data and an expected surface label mode; and according to the output result, allocating a face label mode corresponding to the output result for the face label application user.

In a second aspect, an embodiment of the present application provides a face-to-face label type distribution device, including: the data acquisition module is used for responding to the received surface label application message, and obtaining model input data according to the acquired user basic data and service related data of the surface label application user, wherein the surface label application message is sent by the service provider device in response to the surface label application of the user terminal of the surface label application user; the data processing module is used for inputting model input data into a surface label decision model obtained through pre-training to obtain an output result corresponding to a surface label applying user, the output result is used for representing a surface label mode applicable to the surface label applying user, the surface label mode comprises an upper surface label or a lower surface label, the surface label decision model is obtained through training by utilizing a multi-dimensional classification algorithm according to first sample data of a plurality of sample users, and the first sample data comprises sample user basic data, sample service related data and an expected surface label mode; and the surface label mode allocation module is used for allocating the surface label mode corresponding to the output result for the surface label application user according to the output result.

In a third aspect, an embodiment of the present application provides a face-to-face label mode distribution apparatus, including: a processor and a memory storing computer program instructions; the processor, when executing the computer program instructions, implements the face-to-face assignment method of the first aspect.

In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, where computer program instructions are stored on the computer-readable storage medium, and when the computer program instructions are executed by a processor, the method for allocating a face-to-face label in the first aspect is implemented.

The embodiment of the application provides a method, a device, equipment and a storage medium for allocating a surface signing mode, which can utilize user basic data, service related data and a surface signing decision model of a surface signing application user to enable the surface signing decision model to output an output result which can be used for representing the surface signing mode applicable to the surface signing application user. And allocating the upper line label or the lower line label for the surface label application user according to the output result. The surface label decision model is obtained by training through a multi-dimensional classification algorithm according to the first sample data of a plurality of sample users, so that the automation of a surface label distribution mode for the surface label application users can be realized by utilizing the surface label decision mode, the time spent by the users for surface label application is shortened, and the surface label application efficiency is improved.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the embodiments of the present application will be briefly described below, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.

Fig. 1 is a schematic diagram of an example of an application scenario of a face-to-face label allocation method according to an embodiment of the present application;

fig. 2 is a flowchart of an embodiment of a method for assigning a face-to-face label mode provided in the present application;

fig. 3 is a flowchart of another embodiment of a method for assigning a face-to-face label mode provided in the present application;

fig. 4 is a flowchart of a method for assigning a face-to-face label according to another embodiment of the present disclosure;

fig. 5 is a flowchart of a method for assigning face labels according to another embodiment of the present disclosure;

fig. 6 is a flowchart of an embodiment of a label-on-surface type dispensing device provided in the present application;

fig. 7 is a schematic structural diagram of another embodiment of a label-on-a-surface type dispensing device provided in the present application;

fig. 8 is a schematic structural diagram of a further embodiment of a label-based dispensing device provided in the present application;

fig. 9 is a schematic structural diagram of a further embodiment of a label-on-surface type dispensing device provided in the present application;

fig. 10 is a schematic structural diagram of an embodiment of a label-on-surface type dispensing apparatus provided in the present application.

Detailed Description

Features and exemplary embodiments of various aspects of the present application will be described in detail below, and in order to make objects, technical solutions and advantages of the present application more apparent, the present application will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are intended to be illustrative only and are not intended to be limiting. It will be apparent to one skilled in the art that the present application may be practiced without some of these specific details. The following description of the embodiments is merely intended to provide a better understanding of the present application by illustrating examples thereof.

With the development of internet technology and financial technology, it is difficult for traditional offline business transaction to meet the service requirements of customers, so online business transaction becomes a great development trend. Some financial business needs surface labels, which can be classified into an on-line label and an off-line label. The online surface label can be a remote surface label realized by using an audio and video communication technology. The under-line facing sign is a facing sign that requires the user to go to the financial institution personally. In the process of applying for surface labels, a user needs to select surface label on line or surface label under line according to manual customer guidance or page guidance selection and other modes. But the user operation is more tedious in the process, the time is longer, and the efficiency of the surface label application is lower.

The embodiment of the application provides a method, a device, equipment and a storage medium for allocating a surface-to-label mode, which can acquire user basic data and service related data of a user under the condition that the user applies for the surface-to-label, and automatically acquire a surface-to-label mode judgment score of the user by using the user basic data and the service related data of the user and a surface-to-label decision model obtained by pre-training. The specific surface signing mode can be automatically allocated to the user according to the surface signing mode judging score, so that the operation of the user is reduced, the time required by the surface signing application process is shortened, and the surface signing application efficiency is improved.

The surface signing mode distribution method in the embodiment of the application relates to a user terminal, service provider equipment and surface signing mode distribution equipment. Fig. 1 is a schematic diagram of an example of an application scenario of a face-to-face mode assignment method provided in an embodiment of the present application. As shown in fig. 1, the user terminal 11 may initiate a facebook request to the service provider device 12, and the service provider device 12 may provide data to the facebook allocation device 13.

The user terminal 11 can be used as a channel side to provide service entrance for the user. The user terminal 11 may be used for outputting data and displaying a page to a user, managing a user operation flow, performing user guidance, and the like, but is not limited thereto. User terminal 11 may initiate a request for a facebook application to service provider device 12. The user terminal of the user may include a mobile phone, a computer, a tablet computer, etc., but is not limited thereto. In some examples, the user may also submit a label application through other means, such as a telephone or a location to the service provider, and the like, without limitation.

The service provider may be a business organization such as a bank that can provide business services for the user 11. The service provider device 12 is a device of a service provider, and can receive and process an application from the user terminal 11, and perform qualification verification and other information processing operations. The service provider device 12 may provide data to the face-tag style assignment device 13. The service provider device may be a server or other device, and is not limited herein.

The surface-to-label mode distribution equipment 13 can undertake the responsibility of information transfer and comprehensive verification of the service. And the surface signing application message initiated by the service provider can be received, and the surface signing mode is distributed to the user according to the data provided by the service provider equipment 12, the data stored in advance and the surface signing decision model obtained by pre-training, so that the processes of surface signing application user intention confirmation, identity verification, surface signing authorization and the like are completed.

The specific contents of the method, apparatus, device and storage medium for allocating the face-to-face mode are described in turn below.

The first aspect of the present application provides a face-to-face label manner distribution method, which may be applied to a face-to-face label manner distribution device or a face-to-face label manner distribution apparatus, that is, the face-to-face label manner distribution method may be executed by the face-to-face label manner distribution device or the face-to-face label manner distribution apparatus. Fig. 2 is a flowchart of an embodiment of a method for assigning a face-to-face label mode provided in the present application. As shown in fig. 2, the method for assigning the face-to-face labels may include steps S201 to S203.

In step S201, in response to the received countersign application message, model input data is obtained according to the acquired user basic data and service related data of the countersign application user.

The facebook application message is sent by the service provider device in response to a facebook application from a user terminal of a facebook application user. For the content related to the service provider, the service provider device, the user terminal, and the like, reference may be made to the above related description, and details are not repeated here.

The surface label application user is a user who provides the surface label application. The user basic data is the basic personal information data of the user. The service-related data is information data of the user related to the service. In some examples, the user basic data and the business related data may be provided by the service provider device and the tab mode assigning device together, but are not limited thereto, and may be provided by other devices, such as a device of a credit investigation institution.

After the user basic data and the service related data of the tag application user are obtained, the user basic data and the service related data can be cleaned, and invalid data in the user basic data and the service related data can be screened out, so that the accuracy and the reliability of the user basic data and the service related data are improved.

The service provider device and other devices can generate batch files from the user basic data and the service related data, and transmit the batch files to the surface label mode distribution device in a batch file mode so that the surface label mode distribution device can obtain the user basic data and the service related data. In some examples, the service provider device and other devices may encrypt and desensitize the user basic data and the service related data, and transmit the encrypted and desensitized user basic data and service related data to the profile allocation device through a dedicated line encrypted by a File Transfer Protocol (FTP) to ensure security of the user basic data and the service related data.

In some examples, user profile and business related data may be entered directly into the surface-to-surface decision model as model input data.

In other examples, the cleaned user profile and business related data may be entered as model input data into the face-to-face decision model.

In still other examples, the user profile and the business related data may be classified according to a preset classification rule set, and the model input data may be generated according to the classified class of the user profile and the business related data, and the surface label decision model may be input. The set of classification rules includes a plurality of classification rules for classifying the user profile and the traffic related data. The type of classification rule is not limited herein. Some classification rules are common rules and can consist of three parts if, then, or else; some classification rules are loop rules, which may include a number of common rules that are constructed if, then, or else with loop iteration relationships. For example, common rules include that if a user's age is less than 30 years old, then the user is classified in a first category, otherwise the user is classified in a second category. For another example, the circulation rule includes that if the user age of the user is less than 30 years old, the user is classified into a first category, otherwise, whether the credit level indicated by the user credit investigation information of the user is a high level is judged, if the credit level indicated by the user credit investigation information of the user is a high level, the user is classified into a second category, otherwise, the user is classified into a third category.

In some cases, the categories indicated by the classification rules have attribute labels. The surface label mode distribution equipment can also obtain the user portrait of the surface label application user according to the attribute labels of the user basic data and the category where the service related data is located. The user representation of the user of the face-to-face application may include attribute tags of categories in which the user profile and the business-related data of the user of the face-to-face application are located. The surface label distribution equipment can acquire the problems matched with the user portrait from a preset questionnaire question bank according to the user portrait and generate a business questionnaire matched with the surface label application user. The business questionnaire matched with the face-tag application user includes questions matched with the user representation. The service questionnaire may be provided to the label-applying user during the label-applying process.

For example, the user basic data includes user age and user gender, and the service-related data includes user credit information. Classifying the users with the age less than 40 years into one class according to the classification rule, wherein the corresponding attribute labels are the middle-aged and young-aged people; the classification rule classifies the age of the user to be more than or equal to 40 years old into one class, and the corresponding attribute label is the middle-aged and the old. The classification rule classifies the gender of the user as female, and the corresponding attribute label is female; the classification rule classifies the gender of the user as male, and the corresponding attribute label is male. The classification rule classifies the credit level indicated by the credit investigation information of the user as a high credit level, and the corresponding attribute label is that the credit reaches the standard; the classification rule classifies the credit level indicated by the credit investigation information of the user as a low credit level, and the corresponding attribute label indicates that the credit does not reach the standard. The age of the user A is 41 years old, the sex of the user is female, the credit level indicated by the credit information of the user is high credit level, and the portrait of the user A is a female of middle and old age with the credit reaching the standard. A business questionnaire can be generated for user a that matches the middle aged and elderly women whose credit meets the standard.

In some examples, the user profile may include, but is not limited to, one or more of the following: the system comprises a user card number, a user name, a user gender, a user age, a user certificate number, a user birth date, user occupation information and user residence information.

In some examples, the business-related data includes, but is not limited to, one or more of the following: the mechanism to which the user card belongs, the consumption time of the user card, the consumption address of the user card, the consumption channel of the user card, the service type of the user and credit information of the user.

For example, in the case that the card used by the user is a bank card, the mechanism to which the user card belongs may be a bank to which the bank card of the user belongs, the consumption time of the user card may be consumption time in which the bank card of the user participates, the consumption address of the user card may be an address in which the bank card of the user participates, the consumption channel of the user card may be a channel through which the bank card of the user participates in consumption, the service type of the user is a type of a service to which the user applies for face-signing, and the credit investigation information of the user is credit investigation information of the user.

In step S202, the model input data is input into the pre-trained surface label decision model, and an output result corresponding to the surface label application user is obtained.

And the surface label decision model is obtained by training the first sample data of a plurality of sample users by using a multi-dimensional classification algorithm.

The first sample data comprises sample user basic data, sample service related data and expected surface label mode. The first sample data is sample data used for training the face-to-face decision model. The sample user basic data is user basic data as a sample, that is, user basic data of the sample user. The sample traffic related data is traffic related data as a sample, i.e., traffic related data of a sample user. The expected surface-to-label mode is the surface-to-label mode selected by the sample user, namely the surface-to-label mode according with the intention of the sample user.

In some examples, the facebook decision model may be trained directly with the first sample data.

In other examples, the sample user basic data and the sample service related data in the first sample data may be classified according to a classification rule set, a sample data main body may be generated according to the classified class of the sample user basic data and the sample service related data, and a face-to-face decision model may be obtained by training using the sample data main body. The classification rule set may be consistent with the classification rule set used for classifying the user basic data and the service related data in the above embodiments, and will not be described herein again.

In some examples, the sample user profile may include, but is not limited to, one or more of the following: the system comprises a sample user card number, a sample user name, a sample user gender, a sample user age, a sample user certificate number, a sample user birth date, sample user occupation information and sample user residence information.

The sample user card number is the user card number as the sample, namely the card number of the sample user, the sample user name is the user name as the sample, namely the name of the sample user, the sample user gender is the user gender as the sample, namely the gender of the sample user, the sample user age is the user age as the sample, namely the age of the sample user, the sample user certificate number is the user certificate number as the sample, namely the certificate number of the sample user, the sample user birth date is the user birth date as the sample, namely the birth date of the sample user, the sample user occupation information is the user occupation information as the sample, namely the occupation information of the sample user, and the sample user residence information is the user residence information as the sample, namely the residence information of the sample user.

In some examples, the sample traffic-related data may include, but is not limited to, one or more of the following: the system comprises an organization to which a sample user card belongs, consumption time of the sample user card, a consumption address of the sample user card, a consumption channel of the sample user card, a business type of the sample user and credit investigation information of the sample user.

The sample user card belongs to the mechanism as the sample user card, namely the mechanism of the sample user card, the sample user card consumption time is the sample user card consumption time, namely the sample user card consumption time, the sample user card consumption address is the sample user card consumption address, namely the sample user card consumption address, the sample user card consumption channel is the sample user card consumption channel, namely the sample user card consumption channel, the card sample user service type is the sample user service type, namely the type of the sample user card application business, and the sample user credit information is the sample user credit investigation information, namely the sample user credit investigation information.

Each user basic data can be regarded as one dimension of data, and each service related data can be regarded as one dimension of data. The multidimensional classification algorithm is an algorithm for classifying data with different dimensions, and is not limited herein. For example, the multidimensional classification algorithm may include a bayesian decision algorithm, an XGBOOST algorithm, a decision tree algorithm, a random forest algorithm, and the like, but is not limited thereto.

The model input data is input into a surface label decision model, and the surface label decision model outputs an output result corresponding to the model input data. And the output result is used for representing the face label mode used by the face label application user. The surface label mode comprises an upper thread label or a lower thread label.

In some examples, the output result may include a label-on-label approach or an identification characterizing the label-on-label approach. For example, the output result is 0, which indicates a label on the line; the output result is 1, indicating a label under the line.

In other examples, the output may include a face-to-face sign mode decision score that may characterize the fitness of a face-to-face sign or a face-to-face sign application user, and may also characterize, to some extent, the propensity of the face-to-face sign user to face-to-face sign or face-to-face sign.

In step S203, a face-signing mode corresponding to the output result is allocated to the face-signing applying user according to the output result.

And the output result represents that the surface label applying user is more suitable for the surface label on the line, and the surface label applying user is distributed with the surface label on the line. And the output result represents that the surface label applying user is more suitable for the surface label under the line, and the surface label applying user is distributed with the surface label under the line.

In the embodiment of the application, the user basic data, the service related data and the surface label decision model of the surface label application user can be utilized, so that the surface label decision model outputs an output result which can be used for representing a surface label mode applicable to the surface label application user. And allocating the upper line label or the lower line label for the surface label application user according to the output result. The surface label decision model is obtained by training through a multi-dimensional classification algorithm according to the first sample data of a plurality of sample users, so that the automation of a surface label distribution mode for the surface label application users can be realized by utilizing the surface label decision mode, the time spent by the users for surface label application is shortened, and the surface label application efficiency is improved. In addition, in the embodiment of the application, the user does not need to select the on-line sign or the off-line sign through modes of manual customer guidance, page selection and the like, so that the user operation is simplified, and the user experience is improved.

In some embodiments, outputting the result may include determining the score in a face-to-face manner. Fig. 3 is a flowchart of another embodiment of a method for assigning a face-to-face label mode provided in the present application. Fig. 3 is different from fig. 2 in that step S203 in fig. 2 can be specifically subdivided into step S2031 and step S2032 in fig. 3.

In step S2031, if the countersignature method determination score is within the first threshold range, the user is assigned a line countersignature for the countersignature application.

In step S2032, if the face-tag method determination score is within the second threshold range, a line-under-tag is assigned to the user who applies for the face-tag.

The intersection of the first threshold range and the second threshold range is an empty set. The size and position relationship between the first threshold range and the second threshold range is not limited in the embodiment of the present application.

In some examples, the upper value of the first threshold range may be less than the lower value of the second threshold range, i.e., the face-to-face decision score within the first threshold range is less than the face-to-face decision score within the second threshold range. For example, a target threshold may be set, where the first threshold range is a range smaller than the target threshold, and the second threshold range is a range equal to or larger than the target threshold, that is, if the score is determined to be smaller than the target threshold by the facebook method, a top sign of a line is allocated to the user applying for the facebook, and if the score is determined to be greater than or equal to the target threshold by the facebook method, a bottom sign of the line is allocated to the user applying for the facebook.

In other examples, the lower limit of the first threshold range may be greater than the upper limit of the second threshold range, i.e., the facial pick style determination score within the first threshold range is greater than the facial pick style determination score within the second threshold range. For example, a target threshold may be set, the first threshold range is a range greater than or equal to the target threshold, and the second threshold range is a range smaller than the target threshold, that is, if the score is determined to be smaller than the target threshold by the facebook method, a lower face label is allocated to the user applying for the facebook, and if the score is determined to be greater than or equal to the target threshold by the facebook method, an upper face label is allocated to the user applying for the facebook.

For the convenience of understanding, the example of allocating the top label of the line or the bottom label of the line to the user for applying the top label is described here by taking the example that the top label decision model is obtained by training with a decision tree algorithm, and the lower limit of the first threshold range may be greater than the upper limit of the second threshold range. The algorithm for generating the decision tree may include, but is not limited to, ID3 algorithm, C4.5 algorithm, C5.0 algorithm, and the like.

Each node of the decision tree identifies a class of data, each branch may represent the output of a classification decision, and the leaf nodes of the branches represent the classification results. In the embodiment of the application, the decision results obtained by using the decision tree are two types of label on line or label under line. For example, a first-level node of the decision tree may use the age of the user in the user basic information as a basis for classification, a second-level node of the decision tree may use the professional information of the user in the user basic information as a basis for classification, and a third-level node of the decision tree may use the credit information of the user in the service-related data as a basis for classification; dividing the data into first-layer nodes of which the ages of the users are between 30 and 50, wherein user occupation information represents second-layer nodes of professions with rich economic activities such as entrepreneurs and stockholders, and a face sign mode judgment score corresponding to a face sign application user of a third-layer node with high credit representation information representation credit of the user is located in a first threshold range, and distributing line face signs for the face sign application user; and dividing the data into a first-layer node of which the user age is below 20 years old, representing the user's job practice information as a second-layer node of a simple job which has relatively stable social-economic relations such as a just-entered student, a teacher and a officer, representing the third-layer node in which the user credit information represents the credit, and determining that the mark is located in a second threshold range according to the mark applying mode corresponding to the third-layer node, and distributing marks under the line for the mark applying user. For another example, the first-level node of the decision tree may use the user service type in the service-related data as a basis for classification, where the user service type is a face sign application user of the class I account opening service, and a line lower sign is allocated to the face sign application user.

In the above embodiment, before the face-to-face decision model is used to assign the face-to-face mode to the face-to-face application user, the face-to-face decision model needs to be obtained through training. Fig. 4 is a flowchart of a method for assigning a face-to-face label mode according to another embodiment of the present disclosure. Fig. 4 is different from fig. 2 in that the method for assigning a label pattern shown in fig. 4 may further include step S204 and step S205.

In step S204, first sample data of a plurality of sample users is acquired.

The first sample data is sample data used for training the surface label decision model. The first sample data may be obtained from a data provider. Specifically, the device of the data provider may upload data to the device in the facebook manner as the first sample data. The data provider may include, without limitation, a service provider, other organization, and the like.

In step S205, based on the first sample data, a surface-to-label decision model is trained by using a multidimensional classification algorithm.

And model parameters of the trained surface label decision model enable the probability of the first event to occur to be maximum. The first event comprises a first sample data input surface-to-label decision model, and the surface-to-label mode corresponding to the surface-to-label mode decision score output by the surface-to-label decision model is consistent with the expected surface-to-label mode. The first sample data comprises various multidimensional data, and the surface-to-label decision model obtained through training can output the surface-to-label mode corresponding to input under the action of multidimensional factors.

In some examples, the first sample data for a plurality of sample users may be obtained from a single data provider.

In other examples, the first sample data of the plurality of sample users may also be obtained from N data providers, N being an integer greater than 1.

Specifically, first target sample data of a target sample user can be screened from first sample data of a plurality of sample users provided by each data provider; training to obtain an intermediate model by utilizing a multi-dimensional classification algorithm aiming at first target sample data of a target sample user provided by each data provider; aggregating the model parameters of the intermediate model corresponding to each data provider to obtain integrated model parameters; and updating the intermediate model by using the parameters of the integrated model to obtain a surface label decision model.

In order to obtain a universal and highly accurate countersigning decision model, the countersigning decision model can be obtained by utilizing first sample data of sample users provided by a plurality of data providers. For the convenience of learning, at least part of data of at least part of users can be screened out from the first sample data for model training, but the situation that the data is not screened out after the first sample data is screened out is not excluded. The target sample user is a user selected from the sample users. The first target sample data is first sample data screened from the first sample data. How to filter can be determined according to the relationship of the sample users provided by different data providers and the relationship of the first sample data, and is not determined here.

In some examples, where the coincidence rate of the sample users provided by the N data providers is lower than or equal to a first coincidence rate threshold, and the coincidence rate of the categories of the first sample data provided by the N data providers is higher than or equal to a second coincidence rate threshold, the target sample users include a first category sample user and a second category target sample user, the first target sample data including the first category sample data.

The first category of sample users includes the same sample users provided by N data providers. The second category of sample users includes different sample users provided by N data providers. For example, if N is 2, the 1 st data provider provides the first sample data of the sample user a and the first sample data of the sample user B, and the 2 nd data provider provides the first sample data of the sample user a and the first sample data of the sample user C, the sample user a is the first type of sample user, and the sample user B and the sample user C are the second type of sample user.

The first type of sample data includes first sample data of the same kind provided by the N data providers. The second type of sample data includes first sample data of different types provided by the N data providers. For example, if N is 2, the 1 st data provider provides the sample user age and the sample user occupation information of the sample user, and the 2 nd data provider provides the sample user age and the organization to which the sample user card belongs of the sample user, the sample user age is the first type of sample data, and the sample user occupation information and the organization to which the sample user card belongs are the second type of sample data.

The first coincidence rate threshold is a threshold for determining whether the coincidence degree of the sample users provided by the N data providers is low, and may be determined according to a scene, a demand, experience, and the like, and is not limited herein. The second overlapping rate threshold is a threshold for determining whether the overlapping rate of the first sample data provided by the N data providers is high, and may be determined according to a scene, a demand, experience, and the like, and is not limited herein. The coincidence rate of the sample users provided by the N data providers is lower than or equal to a first coincidence rate threshold value, and the coincidence rate of the categories of the first sample data provided by the N data providers is higher than or equal to a second coincidence rate threshold value, which means that the coincidence rate of the sample users provided by the N data providers is low and the coincidence rate of the categories of the first sample data provided by the N data providers is high. For example, the business features of the sample user groups of the same bank in different regions are similar, i.e. the kind of the first sample data is similar, but the sample users are different. In this case, the first sample data can be screened according to the user dimension, and the first sample data with the same type and the same sample user is selected to participate in the training of the face-to-face decision model.

Specifically, the intermediate model corresponding to each data provider can be trained by using a multidimensional classification algorithm according to the first target sample data of the target sample user provided by each data provider. The model parameters of different intermediate models may be different. The model parameters of the intermediate model corresponding to each data provider can be obtained, and the model parameters of the intermediate model corresponding to each data provider are aggregated to obtain aggregated model parameters. And updating the intermediate model corresponding to each data provider by using the aggregation model parameters, wherein the updated intermediate model can be used as a surface label decision model.

In some cases, each data provider can download the common model from the surface label mode distribution equipment, and each data provider obtains the model parameters of the corresponding intermediate model by using a multidimensional classification algorithm according to the first target sample data of the target sample user provided by the data provider. And each data provider uploads the encrypted gradient of the model parameter of the corresponding intermediate model to the countersigning mode distribution equipment, the countersigning mode distribution equipment aggregates the model parameters of the intermediate model corresponding to each data provider to obtain an aggregated model parameter, and the aggregated model parameter is issued to each data provider so that the data provider updates the intermediate model to obtain the countersigning decision model.

In other cases, the surface signing mode allocation device can train to obtain an intermediate model corresponding to each data provider by using a multidimensional classification algorithm according to first target sample data of a target sample user provided by each data provider, aggregate model parameters of the intermediate model corresponding to each data provider to obtain aggregate model parameters, and update the intermediate model by using the aggregate model parameters to obtain the surface signing decision model.

In other examples, the target sample users include a first type of sample user, and the first target sample data includes a first type of sample data and a second type of sample data, where the coincidence rates of the sample users provided by the N data providers are higher than a third coincidence rate threshold, and the coincidence rates of the types of the first sample data provided by the N data providers are lower than a fourth coincidence rate threshold.

For specific contents of the first type sample user, the first type sample data, and the second type sample data, reference may be made to the relevant description in the above example, and details are not repeated here.

The third threshold of the coincidence rate is a threshold for determining whether the coincidence rate of the sample users provided by the N data providers is high, and may be determined according to a scene, a requirement, experience, and the like, and is not limited herein. The third overlap ratio threshold may be the same as or different from the first overlap ratio threshold, and is not limited herein. The fourth overlapping rate threshold is a threshold for determining whether the overlapping rate of the first sample data provided by the N data providers is low, and may be determined according to a scene, a requirement, experience, and the like, and is not limited herein. The fourth overlap ratio threshold may be the same as or different from the second overlap ratio threshold, and is not limited herein. The coincidence rate of the sample users provided by the N data providers is higher than the third coincidence rate threshold, and the coincidence rate of the types of the first sample data provided by the N data providers is lower than the fourth coincidence rate threshold, indicating that the coincidence rate of the sample users provided by the N data providers is high and the coincidence rate of the types of the first sample data provided by the N data providers is low. For example, the same sample user holds multiple bank credit cards, and the business features of the credit cards of different banks are different, i.e. the kind of the first sample data is different, but the sample users are the same. In this case, the first sample data may be screened according to the dimension of the type of the first sample data, and the first sample data with the same sample user and the different type of the first sample data is selected to participate in the training of the face-to-face decision model.

The first type sample data and the second type sample data of the first type sample user provided by each data provider can be aggregated, and a multi-dimensional algorithm is utilized to train and obtain the surface label decision model.

In some embodiments, in the case that the countersign application user is assigned with the countersign on the line, the countersign application user may also be subjected to risk assessment, so as to further assign an appropriate countersign service mode to the countersign application user. Fig. 5 is a flowchart of a method for assigning a face-to-face label according to another embodiment of the present disclosure. Fig. 5 is different from fig. 2 in that the method for assigning a label on a surface shown in fig. 5 may further include step S206 and step S207.

In step S206, in the case of assigning the top sign of the line to the top sign application user, the user basic data and the business-related data are input into the risk rating model to obtain the risk level of the top sign application user.

The risk level of the surface label applying user is used for indicating the risk level of the surface label applying user. The risk rating model may be trained based on second sample data of the plurality of sample users. The second sample data includes sample user profile data, sample business related data, and known risk levels. For specific contents of the sample user basic data and the sample service related data, reference may be made to the relevant description in the foregoing embodiments, and details are not repeated herein. The known risk level is the risk level that the sample user has determined. And inputting the user basic data and the business related data of the surface signing application user into a risk evaluation model obtained by training second sample data of the sample user, wherein the risk evaluation model can output a risk grade corresponding to the surface signing application user. The training method of the risk rating model may refer to the training method of the surface label decision model in the above embodiments, and details are not repeated herein.

In step S207, a face-to-face service mode is allocated to the face-to-face application user according to the risk level of the face-to-face application user.

The face-to-face service mode indicates that face-to-face service is performed by self-service machine customer service and/or manual customer service. The safety and reliability of the manual customer service are higher than those of the self-service machine customer service. The higher the risk indicated by the risk level of the label applying user is, the higher the security and reliability of the label applying user needs to be allocated with a label service mode.

In some examples, the risk levels include a first risk level, a second risk level, and a third risk level. The risk of the first risk level characterization is higher than the risk of the second risk level characterization, and the risk of the second risk level characterization is higher than the risk of the third risk level characterization.

And indicating that the manual customer service performs the face-to-face service by the face-to-face service mode corresponding to the first risk level. And the surface signing service mode corresponding to the second risk level indicates that the self-service machine customer service and the manual customer service perform surface signing service. And indicating the self-service machine customer service to perform the surface signing service by the surface signing service mode corresponding to the third risk level.

For example, a surface sign application user performs credit card application and card opening business, a surface sign is distributed to the surface sign application user through a surface sign decision model, self-service machine customer service and manual customer service are distributed to the surface sign application user through a risk rating model to perform surface sign service together, and the surface sign service is verified through the manual customer service and assisted by the self-service machine customer service. The self-service machine customer service and the surface label application user carry out conversation, inquire the problem of the surface label application user and automatically feed back; the method comprises the following steps that an artificial customer service participates in the inquiry process of a self-service machine customer service, the online state of a face sign application user of a video call window is observed in real time, and the artificial customer service is switched to inquire in necessary links; and carrying out electronic protocol signing on the video call window, recording the signing track of the face signing application user, and checking and confirming the result in the video call window by manual customer service. The credit card application and card opening business has no breakpoint in the whole process, and the face signing efficiency of the face signing application user is improved, so that the efficiency of the whole business process is improved.

For another example, the surface signing application user performs loan transaction, the surface signing application user is allocated with a line surface signing through a surface signing decision model, and the surface signing application user is allocated with self-service machine customer service and manual customer service through a risk rating model to perform surface signing service together. Before the surface sign application user carries out surface sign, bank approval loan is passed, and the surface sign application user can scan the dynamic two-dimensional code by using the user terminal to enter a surface sign small program for surface sign. And (5) the self-service machine services to check and the manual service to assist in confirmation. The self-service machine customer service and surface sign application user inquires and feeds back the basic service, and after the inquiry is successfully passed, electronic protocol signing is carried out; and (5) manually intervening in the electronic signing process to assist the auditing result.

For example, the surface sign application user performs financial purchasing business, the surface sign application user is allocated with a surface sign through a surface sign decision model, and the surface sign application user is allocated with self-service machine customer service through a risk rating model and is audited by the self-service machine customer service. The self-service machine customer service and the surface sign application user carry out identity verification and risk assessment inquiry. And if the communication failure times of the self-service machine customer service and the face sign application user exceed a communication failure threshold value in the video interaction stage, switching to manual customer service access.

The service of the high-risk face-signing application user is subjected to face signing by the manual customer service, so that the risk can be avoided, and the safety and the reliability of the face signing are improved. The service of the medium-risk face-sign application user is subjected to face-sign by the self-service machine customer service and the manual customer service together, and the manual customer service can timely terminate the face-sign process when finding the risk so as to improve the safety and reliability of the face-sign. The service of the low-risk face-signing application user is subjected to face signing by the self-service machine customer service, so that human resources can be released, and the face signing process is accelerated.

A second aspect of the present application provides a face-tag style dispensing device. Fig. 6 is a flowchart of an embodiment of a label-on-the-surface type dispensing device provided in the present application. As shown in fig. 6, the apparatus 300 for assigning a label-reading style may include a data acquiring module 301, a data processing module 302, and a label-reading style assigning module 303.

The data obtaining module 301 may be configured to obtain model input data according to the obtained user basic data and service related data of the countersign application user in response to the received countersign application message.

The facebook application message is sent by the service provider device in response to a facebook application from a user terminal of a facebook application user.

In some examples, the user profile includes one or more of: the system comprises a user card number, a user name, a user gender, a user age, a user certificate number, a user birth date, user occupation information and user residence information.

In some examples, the traffic-related data includes one or more of: the mechanism to which the user card belongs, the consumption time of the user card, the consumption address of the user card, the consumption channel of the user card, the service type of the user and credit information of the user.

The data processing module 302 may be configured to input the model input data into a pre-trained face-to-face decision model to obtain an output result corresponding to the face-to-face application user.

And the output result is used for representing the face label mode applicable to the face label application user. The surface label mode comprises an upper thread label or a lower thread label. And the surface label decision model is obtained by training the first sample data of a plurality of sample users by using a multi-dimensional classification algorithm. The first sample data comprises sample user basic data, sample service related data and expected surface label mode.

In some examples, the sample user profile includes one or more of: the system comprises a sample user card number, a sample user name, a sample user gender, a sample user age, a sample user certificate number, a sample user birth date, sample user occupation information and sample user residence information.

In some examples, the sample traffic-related data includes one or more of: the system comprises an organization to which a sample user card belongs, consumption time of the sample user card, a consumption address of the sample user card, a consumption channel of the sample user card, a business type of the sample user and credit investigation information of the sample user.

The face-to-face label mode allocation module 303 may be configured to allocate, according to the output result, a face-to-face label mode corresponding to the output result for the face-to-face label application user.

In some examples, the output result includes a face-to-face decision score.

The label-surface-mode assignment module 303 may be configured to: under the condition that the face-to-face sign mode judging score is within a first threshold value range, applying a line face-to-face sign for a face-to-face sign application user; and under the condition that the face-to-face sign mode judgment score is within a second threshold value range, allocating a line lower face sign for the face-to-face sign application user.

The intersection of the first threshold range and the second threshold range is an empty set.

In some examples, the data acquisition module 301 may be configured to: classifying user basic data and service related data according to a preset classification rule set, wherein the classification rule set comprises a plurality of classification rules, and the classification rules are used for classifying the user basic data and the service related data; and generating model input data according to the classified user basic data and the classification of the service related data.

In some embodiments, the categories indicated by the classification rules have attribute labels. Fig. 7 is a schematic structural diagram of another embodiment of a label-on-surface type dispensing device provided in the present application. FIG. 7 differs from FIG. 6 in that the face-tag dispensing apparatus 300 shown in FIG. 7 may further include a representation module 304 and a questionnaire generation module 305.

The portrait module 304 may be configured to obtain a user portrait of the user for whom the face label applies according to the attribute label of the category where the user basic data and the service related data are located.

The questionnaire generation 305 may be used to obtain the question matching with the user portrait from a preset questionnaire question library according to the user portrait, and generate a business questionnaire matching with the face-sign application user.

The business questionnaire matched with the face-tag application user includes questions matched with the user representation.

Fig. 8 is a schematic structural diagram of another embodiment of a label-on-surface type dispensing device provided in the present application. Fig. 8 differs from fig. 6 in that the label-on-the-face dispensing device 300 shown in fig. 8 may further include a sample acquisition module 306 and a training module 307.

The sample acquisition module 306 may be used to acquire first sample data for a plurality of sample users.

The training module 307 may be configured to train to obtain a face-to-face decision model based on the first sample data by using a multidimensional classification algorithm.

And model parameters of the trained surface label decision model enable the probability of the first event to occur to be maximum. The first event comprises a first sample data input surface-to-surface decision model, and the surface-to-surface mode corresponding to the output result output by the surface-to-surface decision model is consistent with the expected surface-to-surface mode.

In some examples, the first sample data of the plurality of sample users is obtained from N data providers, N being an integer greater than 1.

The training module 307 may be configured to filter out first target sample data of a target sample user from the first sample data of a plurality of sample users provided by each data provider; training to obtain an intermediate model by utilizing a multi-dimensional classification algorithm aiming at first target sample data of a target sample user provided by each data provider; aggregating the model parameters of the intermediate model corresponding to each data provider to obtain integrated model parameters; and updating the intermediate model by using the parameters of the integrated model to obtain a surface label decision model.

The first type of sample users comprise the same sample users provided by N data providers. The second category of sample users includes different sample users provided by N data providers. The first type of sample data includes first sample data of the same kind provided by the N data providers. The second type of sample data includes first sample data of different types provided by the N data providers.

Fig. 9 is a schematic structural diagram of a further embodiment of a label-on-surface type dispensing device provided in the present application. Fig. 9 is different from fig. 6 in that the surface-labeling-manner assigning apparatus 300 shown in fig. 9 may further include a risk rating module 308 and a service pattern assigning module 309.

The risk rating module 308 may be configured to input the user basic data and the business related data into the risk rating model to obtain the risk level of the countersign application user when the countersign application user is assigned a line countersign.

The risk level is used for indicating the risk level of the face-pick application user. And the risk rating model is obtained by training according to second sample data of a plurality of sample users. The second sample data includes sample user profile data, sample business related data, and known risk levels.

The mode allocation module 309 may be configured to allocate a face-to-face service mode to the face-to-face application user according to the risk level of the face-to-face application user.

The face-to-face service mode indicates that face-to-face service is performed by self-service machine customer service and/or manual customer service.

A third aspect of the present application provides a face-tag style dispensing device. Fig. 10 is a schematic structural diagram of an embodiment of a label-on-surface type dispensing apparatus provided in the present application. As shown in fig. 10, the face-to-face assignment device 400 includes a memory 401, a processor 402, and a computer program stored on the memory 401 and executable on the processor 402.

In one example, the processor 402 may include a Central Processing Unit (CPU), or an Application Specific Integrated Circuit (ASIC), or may be configured to implement one or more Integrated circuits of the embodiments of the present Application.

Memory 401 may include Read-Only Memory (ROM), Random Access Memory (RAM), magnetic disk storage media devices, optical storage media devices, flash Memory devices, electrical, optical, or other physical/tangible Memory storage devices. Thus, in general, the memory includes one or more tangible (non-transitory) computer-readable storage media (e.g., a memory device) encoded with software comprising computer-executable instructions and when the software is executed (e.g., by one or more processors), it is operable to perform the operations described with reference to the facebook allocation methods in embodiments herein.

The processor 402 runs a computer program corresponding to the executable program code by reading the executable program code stored in the memory 401 for implementing the face-to-face assignment method in the above-described embodiment.

In one example, the face-on-tag style assignment device 400 may also include a communication interface 403 and a bus 404. As shown in fig. 10, the memory 401, the processor 402, and the communication interface 403 are connected by a bus 404 to complete mutual communication.

The communication interface 403 is mainly used for implementing communication between modules, apparatuses, units and/or devices in the embodiments of the present application. Input devices and/or output devices may also be accessed through communication interface 403.

Bus 404 comprises hardware, software, or both that couple the components of the facebook dispensing device 400 to one another. By way of example, and not limitation, Bus 404 may include an Accelerated Graphics Port (AGP) or other Graphics Bus, an Enhanced Industry Standard Architecture (EISA) Bus, a Front-Side Bus (FSB), a HyperTransport (HT) interconnect, an Industry Standard Architecture (ISA) Bus, an InfiniBand interconnect, a Low Pin Count (LPC) Bus, a memory Bus, a Micro Channel Architecture (MCA) Bus, a Peripheral Component Interconnect (PCI) Bus, a PCI-Express (PCI-X) Bus, a Serial Advanced Technology Attachment (SATA) Bus, a Video Electronics Standards Association Local Bus (VLB) Bus, or other suitable Bus, or a combination of two or more of these. Bus 404 may include one or more buses, where appropriate. Although specific buses are described and shown in the embodiments of the application, any suitable buses or interconnects are contemplated by the application.

A fourth aspect of the present application further provides a computer-readable storage medium, where computer program instructions are stored on the computer-readable storage medium, and when the computer program instructions are executed by a processor, the method for allocating a label on a label surface in the foregoing embodiment can be implemented, and the same technical effect can be achieved. The computer-readable storage medium may include a non-transitory computer-readable storage medium, such as a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and the like, which is not limited herein.

It should be clear that the embodiments in this specification are described in a progressive manner, and the same or similar parts in the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments. For apparatus embodiments, device embodiments, computer-readable storage medium embodiments, reference may be made in the descriptive section to method embodiments. The present application is not limited to the particular steps and structures described above and shown in the drawings. Those skilled in the art may make various changes, modifications and additions or change the order between the steps after appreciating the spirit of the present application. Also, a detailed description of known process techniques is omitted herein for the sake of brevity.

Aspects of the present application are described above with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each block of the flowchart illustrations and/or block diagrams, and combinations of blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, enable the implementation of the functions/acts specified in the flowchart and/or block diagram block or blocks. Such a processor may be, but is not limited to, a general purpose processor, a special purpose processor, an application specific processor, or a field programmable logic circuit. It will also be understood that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware for performing the specified functions or acts, or combinations of special purpose hardware and computer instructions.

It will be appreciated by persons skilled in the art that the above embodiments are illustrative and not restrictive. Different features which are present in different embodiments may be combined to advantage. Other variations to the disclosed embodiments can be understood and effected by those skilled in the art upon studying the drawings, the specification, and the claims. In the claims, the term "comprising" does not exclude other means or steps; the word "a" or "an" does not exclude a plurality; the terms "first" and "second" are used to denote a name and not to denote any particular order. Any reference signs in the claims shall not be construed as limiting the scope. The functions of the various parts appearing in the claims may be implemented by a single hardware or software module. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage.

Claims

1. A method for allocating face labels is characterized by comprising the following steps:

responding to a received surface label application message, and obtaining model input data according to the obtained user basic data and service related data of a surface label application user, wherein the surface label application message is sent by a service provider device in response to a surface label application of a user terminal of the surface label application user;

inputting the model input data into a pre-trained surface label decision model to obtain an output result corresponding to the surface label applying user, wherein the output result is used for representing a surface label mode matched with the surface label applying user, the surface label mode comprises an on-line label mode or an off-line label mode, the surface label decision model is obtained by utilizing a multi-dimensional classification algorithm for training according to first sample data of a plurality of sample users, and the first sample data comprises sample user basic data, sample service related data and an expected surface label mode;

and according to the output result, distributing the surface label mode corresponding to the output result for the surface label application user.

2. The method of claim 1, wherein the output result comprises a face-to-face decision score;

according to the output result, distributing a surface label mode corresponding to the output result for the surface label application user comprises the following steps:

under the condition that the face-to-face sign mode judging score is within a first threshold value range, allocating a line face-to-face sign for a face-to-face sign application user;

under the condition that the face-to-face sign mode judging score is within a second threshold value range, allocating a line-to-face sign for the face-to-face sign application user;

wherein an intersection of the first threshold range and the second threshold range is an empty set.

3. The method according to claim 1, wherein the obtaining model input data according to the obtained user basic data and service related data of the user applying for the face-to-face signature comprises:

classifying the user basic data and the service related data according to a preset classification rule set, wherein the classification rule set comprises a plurality of classification rules, and the classification rules are used for classifying the user basic data and the service related data;

and generating the model input data according to the classified user basic data and the classification of the service related data.

4. The method according to claim 1, before obtaining model input data according to the obtained user basic data and service related data of the user applying for the facebook, further comprising:

obtaining the first sample data of a plurality of the sample users;

based on the first sample data, a multi-dimensional classification algorithm is utilized to train and obtain the surface label decision model, model parameters of the surface label decision model obtained through training enable the probability of occurrence of a first event to be maximum, the first event comprises the fact that the first sample data is input into the surface label decision model, and the surface label mode corresponding to the output result output by the surface label decision model is consistent with the expected surface label mode.

5. The method of claim 4, wherein the first sample data of a plurality of the sample users is obtained from N data providers, N being an integer greater than 1;

the training to obtain the face-to-face decision model by utilizing a multi-dimensional classification algorithm based on the first sample data comprises the following steps:

screening first target sample data of a target sample user from the first sample data of a plurality of the sample users provided by each data provider;

training the first target sample data of the target sample user provided by each data provider by utilizing a multi-dimensional classification algorithm to obtain an intermediate model;

aggregating the model parameters of the intermediate model corresponding to each data provider to obtain integrated model parameters;

and updating the intermediate model by using the integrated model parameters to obtain the surface label decision model.

6. The method of claim 5,

under the condition that the coincidence rate of the sample users provided by N data providers is lower than or equal to a first coincidence rate threshold value, and the coincidence rate of the types of the first sample data provided by N data providers is higher than or equal to a second coincidence rate threshold value, the target sample users comprise first type sample users and second type target sample users, and the first target sample data comprises first type sample data;

under the condition that the coincidence rate of the sample users provided by N data providers is higher than a third coincidence rate threshold value and the coincidence rate of the types of the first sample data provided by N data providers is lower than a fourth coincidence rate threshold value, the target sample users comprise the first type of sample data and the second type of sample data;

the first type of sample data comprises the first sample data with the same kind provided by the N data providers, the second type of sample data comprises the different sample users provided by the N data providers, and the second type of sample data comprises the first sample data with the different kinds provided by the N data providers.

7. The method of claim 1, further comprising:

under the condition that the label is distributed on a line for the label-facing application user, inputting the user basic data and the business related data into a risk rating model to obtain a risk grade of the label-facing application user, wherein the risk grade is used for indicating the risk level of the label-facing application user, the risk rating model is obtained by training according to second sample data of a plurality of sample users, and the second sample data comprises the sample user basic data, the sample business related data and known risk grades;

and according to the risk level of the surface signing application user, distributing a surface signing service mode for the surface signing application user, wherein the surface signing service mode indicates that a self-service machine customer service and/or a manual customer service performs surface signing service.

8. The method of claim 7,

the risk levels include a first risk level, a second risk level, and a third risk level, the first risk level characterizing a higher risk than the second risk level characterizing a higher risk than the third risk level characterizing a higher risk;

the surface signing service mode corresponding to the first risk level indicates that surface signing service is carried out by manual customer service, the surface signing service mode corresponding to the second risk level indicates that surface signing service is carried out by self-service machine customer service and manual customer service, and the surface signing service mode corresponding to the third risk level indicates that surface signing service is carried out by self-service machine customer service.

9. The method of claim 3, wherein the classification rule indicates a category having an attribute label;

the method further comprises the following steps:

obtaining a user portrait of the face label application user according to the user basic data and the attribute label of the category where the service related data is located;

and according to the user portrait, obtaining the problems matched with the user portrait in a preset questionnaire question bank, generating a business questionnaire matched with the face sign application user, wherein the business questionnaire matched with the face sign application user comprises the problems matched with the user portrait.

10. The method according to any one of claims 1 to 9,

the user basic data includes one or more of the following: the system comprises a user card number, a user name, a user gender, a user age, a user certificate number, a user birth date, user occupation information and user residence information;

the service related data comprises one or more than two of the following items: the mechanism to which the user card belongs, the consumption time of the user card, the consumption address of the user card, the consumption channel of the user card, the service type of the user and credit information of the user.

11. The method according to any one of claims 1 to 9,

the sample user profile includes one or more of: the method comprises the following steps of (1) obtaining a sample user card number, a sample user name, a sample user gender, a sample user age, a sample user certificate number, a sample user birth date, sample user occupation information and sample user residence information;

the sample business related data comprises one or more than two of the following items: the system comprises an organization to which a sample user card belongs, consumption time of the sample user card, a consumption address of the sample user card, a consumption channel of the sample user card, a business type of the sample user and credit investigation information of the sample user.

12. A face-tag-type dispensing device, comprising:

the data acquisition module is used for responding to the received surface label application message, and obtaining model input data according to the acquired user basic data and service related data of the surface label application user, wherein the surface label application message is sent by the service provider equipment in response to the surface label application of the user terminal of the surface label application user;

the data processing module is used for inputting the model input data into a surface label decision model obtained through pre-training to obtain an output result corresponding to a surface label application user, the output result is used for representing a surface label mode applicable to the surface label application user, the surface label mode comprises a surface label on line or a surface label under line, the surface label decision model is obtained through training by utilizing a multi-dimensional classification algorithm according to first sample data of a plurality of sample users, and the first sample data comprises sample user basic data, sample service related data and an expected surface label mode;

and the surface label mode allocation module is used for allocating the surface label mode corresponding to the output result to the surface label application user according to the output result.

13. A face-tag mode dispensing apparatus, comprising: a processor and a memory storing computer program instructions;

the processor, when executing the computer program instructions, implements the facebook allocation method of any of claims 1-11.

14. A computer-readable storage medium having computer program instructions stored thereon which, when executed by a processor, implement the facebook allocation method of any one of claims 1 to 11.