CN118313921A

CN118313921A - Data processing method, device, equipment, storage medium and product

Info

Publication number: CN118313921A
Application number: CN202410503536.9A
Authority: CN
Inventors: 田媛; 刘一凡; 罗巍; 赵天楷; 张晶晶
Original assignee: China Construction Bank Corp; CCB Finetech Co Ltd
Current assignee: China Construction Bank Corp; CCB Finetech Co Ltd
Filing date: 2024-04-25
Publication date: 2024-07-09

Abstract

The invention discloses a data processing method, a device, equipment, a storage medium and a product, and relates to the technical field of data processing. The method comprises the following steps: carrying out data standardization processing on each initial historical value reference variable of a target object obtained in advance to obtain a corresponding target historical value reference variable; inputting a target historical value reference variable into a pre-established target value prediction model to obtain an object value prediction value corresponding to a target object; acquiring an evaluation reference variable matched with the object liveness of the target object from a pre-established service database; determining an evaluation reference variable box dividing value and an object value box dividing value corresponding to each evaluation reference variable and the object value predicted value; and determining the adjustable field value of the target object according to the evaluation reference variable bin value and the object value bin value, and the pre-acquired object history evaluation value and the fixed field value. The accuracy of the adjustable field value is improved, the actual requirements of users are met, and the user experience is further improved.

Description

Data processing method, device, equipment, storage medium and product

Technical Field

Embodiments of the present invention relate to the field of data processing technologies, and in particular, to a data processing method, apparatus, device, storage medium, and product.

Background

Credit card credit management is an important part of modern credit card management, and credit types are divided into fixed credit, temporary credit and special credit, wherein the business characteristics of the temporary credit comprise the capability of timely meeting the temporary fund demand change of a customer and flexibly responding to market consumption trend, and meanwhile, on scenes and occasions, the temporary card consumption demand of the customer can be timely met, so that the service side is facilitated to enhance the self market competitiveness and increase the income.

The dimension of the user data according to the existing temporary limit adjustment method is single, for example, a service side actively adjusts the temporary limit corresponding to the user only according to the consumption condition of the user and when the user is assessed to meet the temporary limit adjustment condition. Because the dimension of the reference user data is very single, the accuracy of the corresponding determined temporary limit is low, so that a service side is difficult to adjust the temporary limit with a larger amplitude for the user, otherwise, the service side may face a larger risk; and the actual temporary limit requirement of the user cannot be met due to the lower temporary limit, so that the user experience is reduced.

Disclosure of Invention

The embodiment of the invention provides a data processing method, a device, equipment, a storage medium and a product, which are used for solving the problems of low temporary credit accuracy and low user experience caused by single user data dimension referred by the existing temporary credit adjustment method.

In a first aspect, an embodiment of the present invention provides a data processing method, where the method includes:

Carrying out data standardization processing on each initial historical value reference variable of a target object obtained in advance to obtain a corresponding target historical value reference variable;

inputting a target historical value reference variable into a pre-established target value prediction model to obtain an object value prediction value corresponding to a target object; the target value prediction model is a prediction model enabling a value forward prediction value to reach minimum and a value reverse prediction value to reach maximum;

acquiring an evaluation reference variable matched with the object liveness of the target object from a pre-established service database;

Determining an evaluation reference variable box dividing value and an object value box dividing value corresponding to each evaluation reference variable and the object value predicted value;

And determining the adjustable field value of the target object according to the evaluation reference variable bin value and the object value bin value, and the pre-acquired object history evaluation value and the fixed field value.

In a second aspect, an embodiment of the present invention further provides a data processing apparatus, where the apparatus includes:

The data normalization module is used for performing data normalization processing on each initial historical value reference variable of the target object obtained in advance to obtain a corresponding target historical value reference variable;

The object value prediction module is used for inputting the target historical value reference variable into a pre-established target value prediction model to obtain an object value prediction value corresponding to the target object; the target value prediction model is a prediction model enabling a value forward prediction value to reach minimum and a value reverse prediction value to reach maximum;

the variable acquisition module is used for acquiring an evaluation reference variable matched with the object activity of the target object from a pre-established service database;

the box value determining module is used for determining an evaluation reference variable box value and an object value box value corresponding to each evaluation reference variable and the object value predicted value;

and the adjustable field value determining module is used for determining the adjustable field value of the target object according to the evaluation reference variable bin value and the object value bin value, and the pre-acquired object history evaluation value and the fixed field value.

In a third aspect, an embodiment of the present invention further provides an electronic device, including a memory, a processor, and a computer program stored in the memory and capable of running on the processor, where the processor implements a data processing method according to any one of the embodiments of the present invention when the processor executes the program.

In a fourth aspect, embodiments of the present invention further provide a computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements a data processing method according to any of the embodiments of the present invention.

In a fifth aspect, embodiments of the present invention also provide a computer program product comprising a computer program which, when executed by a processor, implements a data processing method according to any of the embodiments of the present invention.

In the embodiment of the invention, through carrying out data standardization processing on each initial historical value reference variable of a pre-acquired target object to obtain a corresponding target historical value reference variable, inputting the target historical value reference variable into a pre-established target value prediction model to obtain an object value predicted value corresponding to the target object, acquiring an evaluation reference variable matched with the object activity of the target object from a pre-established service database, determining an evaluation reference variable box value and an object value box value corresponding to each evaluation reference variable and the object value predicted value, and finally determining an adjustable field value of the target object according to the evaluation reference variable box value and the object value box value, as well as the pre-acquired object historical evaluation value and the fixed field value. By introducing multi-dimensional user data of the target object, namely a target historical value reference characteristic variable, an evaluation reference variable matched with the object liveness, an object historical evaluation value and a fixed field value, the dimension of the referenced user data is more comprehensive, an adjustable field value corresponding to the target object can be accurately determined, the actual temporary quota requirement of the target object is met, and further user experience is improved.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings that are needed in the embodiments will be briefly described below, it being understood that the following drawings only illustrate some embodiments of the present invention and therefore should not be considered as limiting the scope, and other related drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

FIG. 1 is a flow chart of a data processing method provided according to an embodiment of the present invention;

FIG. 2 is a flow chart of another data processing method provided in accordance with an embodiment of the present invention;

FIG. 3 is a schematic diagram of a data processing apparatus according to an embodiment of the present invention;

Fig. 4 is a schematic structural diagram of an electronic device implementing a data processing method according to an embodiment of the present invention.

Detailed Description

In order that those skilled in the art will better understand the present invention, a technical solution in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in which it is apparent that the described embodiments are only some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the present invention without making any inventive effort, shall fall within the scope of the present invention.

It should be noted that the terms "first," "second," and the like in the description and the claims of the present invention and the above figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments of the invention described herein may be implemented in sequences other than those illustrated or otherwise described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.

The technical scheme of the application obtains, stores, uses, processes and the like the data, which all meet the relevant regulations of national laws and regulations.

It should be noted that, in the embodiments of the present application, some existing solutions in the industry such as software, components, models, etc. may be mentioned, and they should be regarded as exemplary, only for illustrating the feasibility of implementing the technical solution of the present application, but it does not mean that the applicant has or must not use the solution.

In an embodiment, fig. 1 is a flowchart of a data processing method according to an embodiment of the present invention, where the method may be implemented by a data processing device, and the data processing device may be implemented in hardware and/or software, and the data processing device may be configured in an electronic device, where the case of determining a corresponding adjustable field value, that is, a temporary credit recommendation, based on multidimensional user data of a target object in a credit card credit management scenario. As shown in fig. 1, the method includes:

S110, performing data standardization processing on each initial historical value reference variable of the target object obtained in advance to obtain a corresponding target historical value reference variable.

The target object may be a user object with consumption potential for a user holding a credit card, that is, the target object may be an adjustable field value to be determined, that is, a temporary limit recommendation frame, and the consumption capability of the target object is reasonably improved by adjusting the adjustable field value, that is, the temporary limit recommendation frame, according with the actual user requirement for the target object, so as to further improve user experience. The initial historical value reference variable may refer to an original historical value reference variable, or may be a historical value reference variable subjected to data preprocessing such as data cleaning, data filtering, data conversion, etc., which is not particularly limited in this embodiment; the historical Value reference feature variable may be understood as a reference feature variable for measuring Customer Value (CMV) of a target object, and exemplary Value reference feature variables may include, but are not limited to: a value forward reference feature variable for measuring revenue conditions, a value reverse reference feature variable for measuring risk loss conditions, etc. In one embodiment, the key components of customer value (CMV) may be pre-analyzed for prediction and assessment of future composite value for different customers, and customer value = benefit-cost, exemplary benefits may include, but are not limited to, normal transaction rebate benefits, recurring credit payouts, staged payouts, annual fees, other revenues, etc.; costs may include, but are not limited to, risk loss, capital costs, operating costs, marketing costs, and the like. The target historical value reference variable may refer to a historical value reference variable obtained after data normalization processing is performed on the initial historical value reference variables, that is, a historical value reference variable after data processing operation in which dimensions and magnitudes among the initial historical value reference variables are kept as consistent as possible. The data normalization process may include at least a Z-score normalization process on the initial historical value reference variables, the Z-score normalization method being a data normalization method for resolving dimensional and magnitude inconsistencies between data.

In an embodiment, an initial historical value reference variable corresponding to a target object may be obtained from a service database of a local or remote server, for example, but not limited to a value forward reference feature variable, a value reverse reference feature variable, and the like, and then data normalization processing such as a Z-score normalization method is performed on the initial historical value reference variable to obtain a corresponding target historical value reference variable. In the actual operation process, in order to eliminate the dimensional influence among different initial historical value reference variables and the influence of the variation size and the numerical value of the variables, a Z-score standardization method can be adopted to carry out data standardization processing on each initial historical value reference variable corresponding to a target object, so as to obtain a corresponding target historical value reference variable, namely, the target historical value reference variable= (initial historical value reference variable-the arithmetic average value of the initial historical value reference variable)/the standard deviation of the initial historical value reference variable, which is obtained after the Z-score standardization processing.

S120, inputting a target historical value reference variable into a pre-established target value prediction model to obtain an object value prediction value corresponding to a target object; the target value prediction model is a prediction model enabling a value forward prediction value to reach minimum and a value reverse prediction value to reach maximum.

The target value prediction model may be a prediction model that minimizes a value forward prediction value and maximizes a value backward prediction value, and the target value prediction model may be used for evaluating an object value prediction value of a target object, and the target value prediction model may be an exemplary convolutional neural network model, that is, a value reference feature variable of the target object is input into a convolutional neural network model trained in advance, so as to obtain a corresponding object value prediction value; in one embodiment, the target value prediction model may include: a value forward prediction model and a value reverse prediction model. The value forward predictor may be a predictor of revenue conditions in the client value of the target object, which may include, but is not limited to: a predicted value of the earnings of the commission, a predicted value of the earnings of the consumption rebate, a predicted value of the earnings of the cyclic interest, etc. The value reverse predictor may be a predictor of risk loss in the client value of the pointer to the target object, and may include, but is not limited to: predicted value of risk loss due to bad account, etc. The object value prediction value may refer to a client value prediction value obtained based on each target historical value reference feature variable of the target object and the target value prediction model.

In the embodiment, a pre-created target value prediction model can be acquired from a storage position such as a local server or a remote server, then a target historical value reference variable obtained after data standardization processing is input into the target value prediction model, and an object value prediction value corresponding to a target object is obtained after model processing. In the actual operation process, the value prediction model may include a value forward prediction model and a value backward prediction model, the target historical value reference variable may include a value forward reference feature variable and a value backward reference feature variable, the value forward reference feature variable and the value backward reference feature variable of the target object are respectively input into the corresponding value forward prediction model and the value backward prediction model, the corresponding value forward prediction value and the corresponding value backward prediction value can be obtained, and then the difference value between the value forward prediction value and the value backward prediction value is used as the object value prediction value corresponding to the target object.

In an embodiment, the value prediction model is a prediction model that makes the value forward prediction value reach minimum and the value reverse prediction value reach maximum, and the object value prediction value corresponding to the target object is equal to the difference value of the value forward prediction value minus the value reverse prediction value, that is, by reducing the specific gravity of the value forward prediction value for measuring the profit situation of the target object and increasing the specific gravity of the value reverse prediction value for measuring the risk loss situation of the target object, the determined object value prediction value of the target object is more accurate, more accords with the actual customer value situation of the target object, and further increases the determination accuracy of the subsequent adjustable field value.

S130, acquiring an evaluation reference variable matched with the object liveness of the target object from a pre-created service database.

The object activity level may be used to measure a business transaction activity condition corresponding to the target object, and the object activity level may be used to determine whether the target object belongs to an active object or an inactive object, for example, the corresponding object activity level may be determined based on the business activity condition of the target object within a preset duration, and the object activity level may be represented in a text (such as high, medium, low) or a number (such as 0-9) or the like.

The evaluation reference variable may be understood as a reference variable for evaluating different service dimensions corresponding to the temporary quota of the target object, and the evaluation reference variable corresponding to the active object and the inactive object may be different or may be partially the same, and exemplary, the evaluation reference variable corresponding to the active object may include, but is not limited to: loan principal balance, card strength floating rating for the past 12 months, average value of usage rate for the past 3 months and the past 6 months, etc.; the evaluation reference variables corresponding to the inactive object may include, but are not limited to: the maximum extra-month transaction amount within the last 12 months, and the last day the address, name or phone number was changed from the date of the day, etc.

In an embodiment, the object activity corresponding to the target object may be determined first, and then the corresponding associated evaluation reference variable may be obtained from the service database of the local or remote server based on the object activity. In the actual operation process, the corresponding object activity degree can be determined according to the service activity condition of the target object within the preset time period, whether the target object belongs to an active object or a non-active object is determined based on the object activity degree, and then an evaluation reference variable matched with the object activity degree of the target object is searched from a service database of a local or remote server.

S140, determining the evaluation reference variable box dividing value and the object value box dividing value corresponding to each evaluation reference variable and the object value predicted value.

The binning method may divide a continuous feature value into a plurality of intervals, where each interval corresponds to a discrete value (binning value), so that feature values in the same interval have similar features, and feature values between different intervals are different as much as possible. The evaluation reference variable binning value may refer to a binning value obtained after the evaluation reference variable is binned by a binning method. The object value binning value may be a binning value obtained by binning the object value prediction value using a binning method.

In the embodiment, a target service sub-box dimension table corresponding to each evaluation reference variable and each target value predicted value can be obtained first, and then the corresponding evaluation reference variable and each target value predicted value are matched by utilizing each target service sub-box dimension table, so that corresponding evaluation reference variable sub-box values and corresponding target value sub-box values are obtained; the target service sub-box dimension table comprises an association mapping relation between an evaluation reference variable or an object value predicted value and a corresponding sub-box value. In the actual operation process, a target box division mode corresponding to each evaluation reference variable and each object value predicted value can be determined, and then the corresponding evaluation reference variable and each object value predicted value are subjected to box division by adopting each target box division mode, so that corresponding evaluation reference variable box division values and object value box division values are obtained.

S150, determining an adjustable field value of the target object according to the evaluation reference variable bin value and the object value bin value, and the pre-acquired object history evaluation value and the fixed field value.

Wherein the object history evaluation value may be understood as the history business evaluation data corresponding to the target object, the object history evaluation value may be distinguished from the value reference feature variable and the evaluation reference variable, and exemplary object history evaluation values may include, but are not limited to: existing credit card customer credit values, existing customer credit values, and the like. The fixed field value may refer to a fixed amount of credit that a target object holding a credit card has, and the fixed field value, i.e., the fixed amount of credit, is not affected by any time constraint, credit rating, and other factors, and the target object may use the amount of credit at any time and any place. The adjustable field value may be understood as a provisional credit recommended frame determined for the target object, the adjustable field values corresponding to the active object and the inactive object may be different, and exemplary, the adjustable field value corresponding to the active object may be higher, the adjustable field value corresponding to the inactive object may be lower, etc.

In an embodiment, the corresponding business feature evaluation value may be determined according to the evaluation reference variable bin value and the object value bin value, and then the corresponding adjustable field value may be determined based on the business feature evaluation value and the object history evaluation value and the fixed field value corresponding to the target object. In the actual operation process, character string combination can be performed on the evaluation reference variable sub-box value and the object value sub-box value to generate a corresponding sub-box value combination sequence number, the sub-box value combination sequence number is used for matching a corresponding target service combination sub-box dimension table to obtain a corresponding service characteristic matching value, and then weighting and summing are performed on all the obtained service characteristic matching values to obtain a corresponding service characteristic evaluation value, wherein the target service combination sub-box dimension table comprises a sub-box value combination sequence number generated after character string combination is performed on the evaluation reference variable and the object value prediction value and an association mapping relation between the corresponding sub-box values; then, respectively classifying the service characteristic evaluation value, the object history evaluation value and the fixed field value according to a classification mode matched with the object activity of the target object to obtain a corresponding service characteristic classification value, an object history classification value and a fixed field value classification value; and finally, respectively matching a preset adjustment coefficient matching dimension table and a basic adjustment value matching dimension table by using the business characteristic box value, the object history box value and the fixed field value box value to obtain a corresponding adjustment coefficient and a basic adjustment value, and taking the product result among the fixed field value, the adjustment coefficient and the basic adjustment value as an adjustable field value corresponding to the target object, wherein the adjustment coefficient matching dimension table comprises an association mapping relation among the business characteristic box value, the fixed field value box value and the corresponding adjustment coefficient, and the basic adjustment value matching dimension table comprises an association mapping relation among the object history box value, the fixed field value box value and the corresponding basic adjustment value.

According to the technical scheme, the data standardization processing is carried out on each initial historical value reference variable of the target object, so that the dimensional influence among different initial historical value reference variables and the influence of the variation size and the numerical value of the variables are eliminated, and the effectiveness and the accuracy of the historical value reference variables are further improved; the corresponding object value predicted value is determined based on the target historical value reference variable and the target value predicted model, and the target value predicted model is a predicted model which enables the value forward predicted value to reach minimum and the value reverse predicted value to reach maximum, so that the determined object value predicted value of the target object is more accurate, the actual client value situation of the target object is more met, and the determination accuracy of the follow-up adjustable field value is further improved; by introducing multi-dimensional user data of the target object, namely a target historical value reference characteristic variable, an evaluation reference variable matched with the object liveness, an object historical evaluation value and a fixed field value, the dimension of the referenced user data is more comprehensive, an adjustable field value corresponding to the target object can be accurately determined, the actual temporary quota requirement of the target object is met, and further user experience is improved.

In one embodiment, before performing data normalization processing on each initial historical value reference variable of the pre-acquired target object to obtain a corresponding target historical value reference variable, the method further includes:

determining a corresponding historical time window according to the variable characteristics of each initial historical value reference variable;

and acquiring an initial historical value reference variable corresponding to the target object in the historical time window.

The variable feature may be understood as a customer value evaluation feature corresponding to an initial historical value reference variable, so that in order to make the selected initial historical value reference variable more conform to the actual customer value situation of the target object, the initial historical value reference variable in different historical time periods (i.e. a historical time window) may be used as an important data source of a subsequent target value prediction model, and by taking an example of a stage revenue element corresponding to the target object, the following initial historical value reference variable may be selected, where the initial historical value reference variable is not limited to: the account-free stage balance of the current month, the average value of the cash interest taken in the past 3 months, the continuously increasing month number of the consumption amount of the last 6 months, the stage income of the last 12 months, and the like.

In an embodiment, variable characteristics of each initial historical value reference variable can be extracted, then a historical time window corresponding to each initial historical value reference variable is determined according to the variable characteristics, and then the initial historical value reference variable corresponding to the target object in the historical time window is obtained from a service database of a local or remote server. The corresponding historical time window is determined according to the variable characteristics of the initial historical value reference variable, and then the initial historical value reference variable corresponding to the target object in the historical time window is obtained, so that the obtained initial historical value reference variable is more in line with the actual client value condition of the target object, and the accuracy of the value predicted value of the target object is effectively improved.

In one embodiment, the training process of the target value prediction model includes:

selecting individual data with adjustable field value using records as original sample data;

randomly extracting partial data from the original sample data as a corresponding training set and a corresponding testing set;

And inputting sample data in the training set into a pre-established original value prediction model for continuous iterative training until the obtained target value prediction model enables a value forward prediction value obtained by testing test data in the testing set to be minimum and a value reverse prediction value to be maximum.

The adjustable field value usage record may be understood as a usage situation that different users recommend a frame for an adjustable field value, that is, a temporary quota, and based on the adjustable field value usage record, raw sample data with more effective and more practical reference meaning may be selected, for example: 2 ten thousand yuan of temporary limit recommendation frames are respectively distributed for the user A and the user B, but the user A only uses 10% of the temporary limit recommendation frames, the user B uses 95%, and obviously the adjustable field value use record corresponding to the user B has a reference meaning and is suitable for being used as original sample data.

In an embodiment, individual data with adjustable field value usage records of different users can be selected from a service database of a local or remote server as original sample data required by a subsequent building model, then a part of sample data is randomly extracted from the original sample data, a corresponding training set and a corresponding testing set are built based on a preset data set segmentation proportion (for example, the training set 7: the testing set 3), and finally the training set and the testing set are utilized to sequentially perform iterative training and testing on the original value prediction model which is built in advance until the obtained target value prediction model enables a value forward prediction value obtained by testing test data in the testing set to be minimum and a value reverse prediction value to be maximum. In the training process of the target value prediction model, the specific gravity of the value forward predicted value for measuring the profit situation of different users is reduced, and the specific gravity of the value reverse predicted value for measuring the risk loss situation of different users is improved, so that the target value predicted value of a target object which is determined later is more accurate, the actual client value situation of the target object is more met, and the accuracy of determining the field value which can be adjusted later is improved.

In an embodiment, fig. 2 is a flowchart of another data processing method according to an embodiment of the present invention, and this embodiment is a further description of a data processing method based on the foregoing embodiment. As shown in fig. 2, the method includes:

S210, performing data standardization processing on each initial historical value reference variable of the target object obtained in advance to obtain a corresponding target historical value reference variable.

S220, inputting the value forward reference characteristic variable into a pre-established value forward prediction model to obtain a corresponding value forward prediction value; the value forward prediction model is a prediction model which minimizes a value forward prediction value.

Wherein the value forward reference feature variable may refer to a reference feature variable for measuring a revenue condition of the target object, and exemplary value forward reference feature variables may include, but are not limited to: a stage commission reference feature variable, a consumption rebate reference feature variable, a cycle interest reference feature variable, and the like.

The value forward prediction model may refer to a prediction model that minimizes a value forward prediction value, and the value forward prediction model may be used to evaluate the value forward prediction value of the target object, and may be, for example, a convolutional neural network model, that is, a value forward reference feature variable of the target object is input into a corresponding convolutional neural network model trained in advance, so as to obtain a corresponding value forward prediction value; in an embodiment, a plurality of historical value forward reference feature variables of users can be used for model training of the corresponding convolutional neural network model, model parameters are continuously optimized, and then a trained value forward prediction model is obtained.

In the embodiment, a value forward reference characteristic variable corresponding to a target object can be obtained from a service database of a local or remote server, and then the value forward reference characteristic variable is input into a pre-constructed value forward prediction model, so that a corresponding value forward prediction value is obtained; the value forward prediction model is a prediction model which minimizes a value forward prediction value.

In one embodiment, the corresponding value forward prediction values may be determined in a value forward prediction model by weighted summing all value forward reference feature variables. Taking the predicted value of the stage commission benefit corresponding to the stage commission reference characteristic variable as an example, assuming that the stage commission reference characteristic variable comprises three business dimensions of the current month unsettled stage balance, the past 12 months monthly stage income and the new basic integral of the current month corresponding to the target object, the process for determining the predicted value of the stage commission benefit is as follows: and carrying out weighted summation on three stage commission reference characteristic variables after data standardization processing (such as Z-score standardization processing), so as to obtain corresponding stage commission income prediction values.

S230, inputting the value reverse reference characteristic variable into a pre-created value reverse prediction model to obtain a corresponding value reverse prediction value; the value reverse prediction model is a prediction model that maximizes a value reverse prediction value.

Wherein the value back reference feature variable may refer to a reference feature variable for measuring risk loss condition of the target object, and exemplary value back reference feature variables may include, but are not limited to: risk reference feature variables due to bad account, etc.

The value reverse prediction model may refer to a prediction model that maximizes a value reverse prediction value, and the value reverse prediction model may be used to evaluate the value reverse prediction value of the target object, and may be, for example, a convolutional neural network model, where a model training process may refer to a training process of the value forward prediction model, which is not described herein.

In an embodiment, a value reverse reference feature variable corresponding to a target object can be obtained from a service database of a local or remote server, and then the value reverse reference feature variable is input into a pre-constructed value reverse prediction model, so that a corresponding value reverse prediction value is obtained; the value reverse prediction model is a prediction model that maximizes a value reverse prediction value.

S240, in the case of containing at least two value forward predicted values, determining a corresponding value forward predicted total value according to the sum of each value forward predicted value, and determining a corresponding object value predicted value according to the difference between the value forward predicted total value and the value reverse predicted value.

In an embodiment, a corresponding object value predicted value determining manner may be selected according to a specific number of the value forward predicted values, and specifically, when the number of the value forward predicted values is at least two, a sum of all the value forward predicted values, that is, a value forward predicted total value, may be determined first, and then a difference between the value forward predicted total value and the value reverse predicted value is taken as an object value predicted value corresponding to the target object. In actual operation, the value forward predictor may include: the stage commission profit prediction value, the consumption rebate profit prediction value, and the cyclic interest profit prediction value, the value reversal prediction value may include: and because of the risk loss predicted value caused by bad account, the object value predicted value corresponding to the target object is equal to the period commission profit predicted value, the consumption rebate profit predicted value, the circulation interest profit predicted value and the risk loss predicted value.

S250, when the value forward predicted value is included, determining a corresponding object value predicted value according to a difference value between the value forward predicted value and the value reverse predicted value.

In an embodiment, when the number of the value forward predictors is only one, the difference between the value forward predictors and the value reverse predictors may be directly taken as the object value predictor corresponding to the target object.

S260, determining the corresponding object activity according to the service activity condition of the target object in the preset time.

The preset duration may refer to a time period that is preconfigured to determine the object activity of the target object, and exemplary, the preset duration may include, but is not limited to: 6 months, 12 months, etc. The service activity condition may refer to various service transaction conditions of the target object within a preset duration, and exemplary service activity conditions may include, but are not limited to: total value of the amount of consumption, amortization, personal information change, etc.

In an embodiment, service data of a target object within a preset duration may be acquired, a corresponding service activity condition is determined based on the service data, and then a corresponding object activity level is determined according to the service activity condition.

S270, determining whether the target object is an active object according to the object liveness and a preconfigured liveness threshold value.

The liveness threshold may be a pre-configured object liveness threshold for determining an active object or an inactive object.

In an embodiment, whether the target object is an active object may be determined according to the determined object liveness and a pre-configured liveness threshold. In the actual operation process, when the object liveness is expressed in a high-medium-low mode, if the object liveness corresponding to the target object is low, determining that the target object is an inactive object; and if the object activity degree corresponding to the target object is medium or high, determining that the target object is an active object.

And S280, when the target object is an active object, searching an active reference variable matched with the active object from a pre-created service database.

Wherein an active reference variable may be understood as an evaluation reference variable corresponding to an active object, exemplary active reference variables may include, but are not limited to: loan principal balance, last 12 months card strength float rating, last 3 months bottom line usage average, last 6 months average stage revenue, etc.

In an embodiment, when the target object is determined to be an active object, the active reference variable corresponding to the association may be obtained from a service database of a local or remote server. In the actual operation process, a data table containing all objects and all active reference variables can be maintained in a service database, so that when the target object is determined to be the active object, the corresponding matched active reference variables can be directly searched in the corresponding data table according to the identification information of the target object.

And S290, searching an inactive reference variable matched with the inactive object from a pre-created service database when the target object is the inactive object.

Wherein, the inactive reference variable may be understood as an evaluation reference variable corresponding to the inactive object, and exemplary inactive reference variables may include, but are not limited to: the maximum extra-month transaction amount within the last 12 months, the last day the address, name or phone number was changed from the date of the day, etc.

In an embodiment, reference may be made to the implementation procedure of S280, and when it is determined that the target object is an inactive object, a corresponding associated inactive reference variable may be obtained from a service database of a local or remote server.

Because of the large difference between the inactive guest group and the active guest group, the embodiment selects the corresponding matched evaluation reference variable according to the object liveness of the target object, the selected data is more effective, the actual situation of the target object is more met, and the accuracy of the value predicted value of the target object is further effectively improved.

S2100, determining a corresponding target box division mode according to the variable types of the evaluation reference variable and the object value predicted value.

Wherein, variable type may refer to the characteristic representation corresponding to the evaluation reference variable and the object value predicted value, and exemplary variable types may include, but are not limited to: text type, numeric type, etc. The target box division manner may refer to a box division manner corresponding to the evaluation reference variable and the object value predicted value, and, by way of example, the target box division manner may at least include a target service box dimension table corresponding to the evaluation reference variable and the object value predicted value, and the like.

In an embodiment, the corresponding target binning mode may be obtained from a storage location such as a local or remote server according to a variable type, e.g., text type, numeric type, etc., that evaluates the reference variable and the object value prediction value. In the actual operation process, corresponding target service box dimension tables can be respectively obtained according to variable types of the evaluation reference variable and the object value predicted value so as to determine evaluation reference variable box values and object value box values corresponding to the evaluation reference variable and the object value predicted value based on the target service box dimension tables. For example, taking a target service box dimension table corresponding to the object value predicted value as an example, the target service box dimension table may include an association mapping relationship between the object value predicted values and the corresponding box values of different intervals, for example, the interval (18, 38) corresponds to the box value of 1, the interval [38,54 ] corresponds to the box value of 2, and the like.

S2110, the target box division mode is adopted to divide each evaluation reference variable and each object value predicted value into boxes, and corresponding evaluation reference variable box division values and object value box division values are obtained.

In an embodiment, the target box-dividing mode corresponding to each evaluation reference variable and each object value predicted value can be utilized, for example, a target service box-dividing dimension table is utilized to carry out box-dividing matching on the corresponding evaluation reference variable and the corresponding object value predicted value, so as to obtain a corresponding evaluation reference variable box-dividing value and an object value box-dividing value.

S2120, determining a corresponding service characteristic evaluation value according to the evaluation reference variable bin value and the object value bin value.

The service characteristic evaluation value can be understood as an evaluation value for measuring the provisional limit amplitude-up condition of the target object.

In an embodiment, the service feature evaluation value corresponding to the target object may be comprehensively determined based on the determined evaluation reference variable bin value and the object value bin value.

In an embodiment, when the target object is an active object, evaluating the reference variable bin value as an active reference variable bin value; determining a corresponding service feature evaluation value according to the evaluation reference variable bin value and the object value bin value, including:

Performing character string combination on the active reference variable bin value and the object value bin value to obtain a corresponding service potential matching value and a first service balance matching value;

And carrying out weighted summation on the service potential matching value and the first service balance matching value to obtain a corresponding service characteristic evaluation value.

Wherein, the active reference variable bin value may refer to an estimated reference variable bin value corresponding to the active object, and exemplary active reference variable bin values may include, but are not limited to: the loan principal balance is divided into a box value, a box value of the last 12 months card strength floating evaluation, a box value of the last 3 months bottom limit use rate average value, a box value of the last 6 months average stage income, and the like. The business potential match value may be a match value of a pointer to an active object to reflect the user's transaction potential, and exemplary business potential match values may include, but are not limited to: consumption promotion matching values, etc. The first business balance match value may be a match value of a pointer to an active object to reflect a user liability balance condition, and exemplary first business balance match values may include, but are not limited to: a staging balance match, a loan balance match, etc.

In an embodiment, when the target object is an active object, the evaluation reference variable bin value may be an active reference variable bin value, the active reference variable bin value and the object value bin value may be combined according to a preset sequencing sequence, meanwhile, a separator 'i' may be used to segment adjacent bin values, so as to obtain a corresponding bin value combination sequence number, and then the bin value combination sequence number is used to match a corresponding target service combination bin dimension table, so as to obtain a corresponding service potential matching value and a first service balance matching value, and finally the service potential matching value and the first service balance matching value are weighted and summed, so as to obtain a corresponding service feature evaluation value. In an actual operation process, taking a consumption promotion matching value determining process in the business potential matching values as an example, the assumption is made that the corresponding active reference variable box division value comprises: the last 12 months use card intensity to float and appraise the box value, and last 3 months month end use rate average value box value, the correspondent box value in the goal business combination box dimension table (active customer group consumption promotes the dimension table) built in advance is the place of the box value combination serial number as follows fixed format: the object value box value is equal to the value box value of the last 12 months card intensity floating rating box value equal to the value box value of the last 3 months month end limit use rate average value, for example, the box value combination sequence number 1-2 corresponds to the associated consumption promotion matching value of-2, and the consumption promotion matching value corresponds to the associated consumption promotion matching value of 1.719 and the like, so that the character string combination is carried out through the object corresponding active reference variable box value and the object value box value, and the generated box value combination sequence number is utilized to match the active guest group consumption promotion dimension table, so that the corresponding consumption promotion matching value can be obtained.

In an embodiment, when the target object is an inactive object, evaluating the reference variable bin value as an inactive reference variable bin value; determining a corresponding service feature evaluation value according to the evaluation reference variable bin value and the object value bin value, including:

Performing character string combination on the inactive reference variable bin value and the object value bin value to obtain a corresponding active rate matching value and a second service balance matching value;

And carrying out weighted summation on the activity rate matching value and the second service balance matching value to obtain a corresponding service characteristic evaluation value.

Wherein, the inactive reference variable bin value may refer to an evaluation reference variable bin value corresponding to the inactive object, and exemplary inactive reference variable bin values may include, but are not limited to: the maximum overseas month transaction amount and address binning value within the last 12 months, and the last time the address, name or phone number has been changed from the today's binning value, etc. The liveness match value may be a match value for the inactive object that reflects the user's transacting liveness. The second business balance match value may be a match value indicating a pointer to an inactive object for reflecting a user liability balance condition, and exemplary second business balance match values may include, but are not limited to: loan balance matching values, etc.

In an embodiment, when the target object is an inactive object, the evaluation reference variable bin value may be an inactive reference variable bin value, the inactive reference variable bin value and the object value bin value may be combined according to a preset sequencing sequence, meanwhile, a separator 'i' may be used to segment adjacent bin values, so as to obtain a corresponding bin value combination sequence number, and then the bin value combination sequence number is used to match a corresponding target service combination bin dimension table, so as to obtain a corresponding active rate matching value and a second service balance matching value, and finally, the active rate matching value and the second service balance matching value are weighted and summed, so as to obtain a corresponding service feature evaluation value.

S2130, classifying the service characteristic evaluation value, the object history evaluation value and the fixed field value according to a classification mode matched with the object activity of the target object to obtain a corresponding service characteristic classification value, an object history classification value and a fixed field value classification value.

The service characteristic binning value, the object history binning value and the fixed field value binning value may refer to a service characteristic evaluation value, an object history evaluation value and a binning value corresponding to a fixed field value, respectively.

In an embodiment, a target service binning dimensional table pre-constructed for a service feature evaluation value, an object history evaluation value and a fixed field value may be obtained according to a binning manner matched with the object activity of the target object, and the corresponding service feature binning value, object history binning value and fixed field value binning value may be obtained by matching the service feature evaluation value, the object history evaluation value and the fixed field value with the corresponding target service binning dimensional table.

S2140, determining the adjustable field value of the target object according to the business feature binning value, the object history binning value and the fixed field value binning value.

In an embodiment, the final adjustable field value may be determined based on the previously determined business feature binning value, object history binning value, and fixed field value binning value. In the actual operation process, the business characteristic box division value, the object history box division value and the fixed field value box division value can be matched with a preset adjustment coefficient matching dimension table and a preset basic adjustment value dimension table to obtain corresponding adjustment coefficients and basic adjustment values, and then the product result among the fixed field value, the adjustment coefficients and the basic adjustment values is used as an adjustable field value corresponding to the target object, namely, the temporary limit recommendation frame.

In one embodiment, determining the adjustable field value of the target object based on the business feature binning value, the object history binning value and the fixed field value binning value comprises:

determining a corresponding adjustment coefficient according to the service characteristic box division value and the fixed field value box division value;

determining a corresponding basic adjustment value according to the object history bin value and the fixed field value bin value;

determining a corresponding adjustable base value according to the base adjustment value and the fixed field value;

An adjustable field value of the target object is determined based on the adjustable base value and the adjustment coefficient.

The adjustment coefficient may refer to a coefficient corresponding to the adjustable field value, i.e., the provisional limit recommendation frame. The basic adjustment value may refer to an adjustable field value, i.e. a basic adjustment value corresponding to the provisional credit recommendation.

In an embodiment, a service characteristic box division value and a fixed field value box division value are adopted to match a pre-constructed adjustment coefficient matching dimension table to obtain a corresponding adjustment coefficient, an object history box division value and a fixed field value box division are adopted to match a pre-constructed basic adjustment value matching dimension table to obtain a corresponding basic adjustment value, then the product of the basic adjustment value and the fixed field value is used as an adjustable basic value corresponding to a target object, and finally the product of the adjustable basic value and the adjustment coefficient is determined as an adjustable field value corresponding to the target object, namely a temporary quota recommendation frame.

In an embodiment, the data processing method provided by the embodiment of the present invention further includes:

acquiring the adjustable field value use condition of the target object aiming at the adjustable field value in the adjustable field value use period;

The target value prediction model is dynamically adjusted based on the adjustable field value usage until the adjustable field value is optimal.

The adjustable field value lifetime may refer to a length of time that the target object recommends a frame using an adjustable field value, i.e., a temporary credit, and exemplary adjustable field value lifetime may include, but is not limited to: 12 months, 24 months, etc. The adjustable field value usage may refer to a usage of the target object for the adjustable field value, i.e. the temporary credit recommended frame, for example, the adjustable field value usage may at least include a temporary credit usage percentage, etc.

In an embodiment, after determining the adjustable field value corresponding to the target object, the adjustable field value usage condition of the adjustable field value of the target object in the service life of the adjustable field value may be tracked, and then the target value prediction model is dynamically adjusted according to the adjustable field value usage condition until the adjustable field value reaches the optimum. In the actual operation process, if the usage percentage of the temporary quota recommended for the adjustable field value, i.e. the temporary quota, by the target object in the service life of the adjustable field value is lower than a preset threshold, the determined adjustable field value is too high, and the temporary quota is wasted, and at this time, the service dimension of the value reference feature variable can be increased in the target value prediction model for the target object, so that the value predicted value of the object determined according to the value reference feature variable of the target object better meets the temporary quota requirement of the target object, and the service dimension of the value reference feature variable can be increased continuously until the adjustable field value reaches the optimum; if the usage percentage of the target object for the temporary quota of the adjustable field value in the lifetime of the adjustable field value is higher than the preset threshold, it is indicated that the determined adjustable field value is lower and may not meet the temporary quota requirement of the target object, and at this time, the service dimension of the value reference feature variable may be reduced in a proper amount in the target value prediction model for the target object, so that the newly determined adjustable field value is optimal. The adjustable field value is adjusted and optimized by utilizing the use condition of the adjustable field value of the target object, so that the adjusted adjustable field value is more in line with the actual user requirement, and further the user experience is effectively improved.

According to the technical scheme, the data standardization processing is carried out on each initial historical value reference variable of the target object, so that the dimensional influence among different initial historical value reference variables and the influence of the variation size and the numerical value of the variables are eliminated, and the effectiveness and the accuracy of the historical value reference variables are further improved; the target object value predicted value is introduced to comprehensively evaluate the client value of the target object, and meanwhile, the evaluation reference variable matched with the target object is selected according to the liveness difference of different guest groups, so that the actual client value situation of the target object is more met, and the determination accuracy of the follow-up adjustable field value is further improved; by adopting the box dividing method to determine the box dividing values of different user data dimensions, the adjustable field values corresponding to the target objects are further determined, the accuracy of determining the adjustable field values is improved, the actual temporary limit requirements of the target objects can be met to the maximum extent, so that the user experience is improved, and meanwhile, the risk of a service side can be considered.

In an embodiment, fig. 3 is a schematic structural diagram of a data processing apparatus according to an embodiment of the present invention. As shown in fig. 3, the apparatus includes: a data normalization module 31, an object value prediction module 32, a variable acquisition module 33, a binning value determination module 34 and an adjustable field value determination module 35.

The data normalization module 31 is configured to perform data normalization processing on each initial historical value reference variable of the target object obtained in advance, so as to obtain a corresponding target historical value reference variable.

The object value prediction module 32 is configured to input a target historical value reference variable into a pre-created target value prediction model to obtain an object value predicted value corresponding to the target object; the target value prediction model is a prediction model enabling a value forward prediction value to reach minimum and a value reverse prediction value to reach maximum.

The variable obtaining module 33 is configured to obtain, from a pre-created service database, an evaluation reference variable that matches the object activity of the target object.

The bin value determining module 34 is configured to determine an estimated reference variable bin value and an object value bin value corresponding to each estimated reference variable and the object value predicted value.

An adjustable field value determining module 35, configured to determine an adjustable field value of the target object according to the evaluation reference variable bin value and the object value bin value, and the object history evaluation value and the fixed field value acquired in advance.

According to the embodiment of the invention, a data standardization module is used for carrying out data standardization processing on each initial historical value reference variable of a target object obtained in advance to obtain a corresponding target historical value reference variable, then the target historical value reference variable is input into a pre-established target value prediction model through an object value prediction module to obtain an object value prediction value corresponding to the target object, a variable obtaining module is used for obtaining an evaluation reference variable matched with the object activity of the target object from a pre-established service database, a box value determining module is used for determining an evaluation reference variable box value and an object value box value corresponding to each evaluation reference variable and the object value prediction value, and finally an adjustable field value determining module is used for determining an adjustable field value of the target object according to the evaluation reference variable box value and the object value box value, and the pre-obtained object historical evaluation value and fixed field value. By introducing multi-dimensional user data of the target object, namely a target historical value reference characteristic variable, an evaluation reference variable matched with the object liveness, an object historical evaluation value and a fixed field value, the dimension of the referenced user data is more comprehensive, an adjustable field value corresponding to the target object can be accurately determined, the actual temporary quota requirement of the target object is met, and further user experience is improved.

In an embodiment, the data processing apparatus further comprises:

the time window determining module is used for determining a corresponding historical time window according to the variable characteristics of each initial historical value reference variable before carrying out data standardization processing on each initial historical value reference variable of a pre-acquired target object to obtain the corresponding target historical value reference variable;

The initial historical value reference variable acquisition module is used for acquiring initial historical value reference variables corresponding to the target object in the historical time window.

In one embodiment, the target value prediction model includes: a value forward prediction model and a value reverse prediction model; the target historical value reference variables include: a value forward reference feature variable and a value reverse reference feature variable; accordingly, the object value prediction module 32 includes:

The value forward prediction value determining unit is used for inputting the value forward reference characteristic variable into a pre-created value forward prediction model to obtain a corresponding value forward prediction value; the value forward prediction model is a prediction model which enables a value forward prediction value to reach minimum;

The value reverse prediction value determining unit is used for inputting the value reverse reference characteristic variable into a pre-created value reverse prediction model to obtain a corresponding value reverse prediction value; the value reverse prediction model is a prediction model which enables a value reverse prediction value to reach the maximum;

A first object value prediction value determining unit, configured to determine, in a case where at least two value forward prediction values are included, a corresponding value forward prediction total value according to a sum of each value forward prediction value, and determine a corresponding object value prediction value according to a difference between the value forward prediction total value and the value reverse prediction value;

And the second object value predicted value determining unit is used for determining a corresponding object value predicted value according to the difference value between the value forward predicted value and the value reverse predicted value under the condition that one value forward predicted value is included.

In one embodiment, evaluating the reference variable includes: active and inactive reference variables; the variable acquisition module 33 includes:

the activity determining unit is used for determining the corresponding object activity according to the service activity condition of the target object within the preset duration;

the active object determining unit is used for determining whether the target object is an active object according to the object activity level and a preconfigured activity level threshold value;

The active reference variable searching unit is used for searching active reference variables matched with the active objects from a pre-created service database when the target objects are the active objects;

And the inactive reference variable searching unit is used for searching the inactive reference variable matched with the inactive object from a pre-created service database when the target object is the inactive object.

In one embodiment, the binning value determination module 34 comprises:

The box division mode determining unit is used for determining a corresponding target box division mode according to the variable types of the evaluation reference variable and the object value predicted value;

The first box dividing value determining unit is used for dividing each evaluation reference variable and each object value predicted value into boxes by adopting a target box dividing mode to obtain corresponding evaluation reference variable box dividing values and object value box dividing values.

In one embodiment, the adjustable field value determination module 35 includes:

The business characteristic evaluation value determining unit is used for determining a corresponding business characteristic evaluation value according to the evaluation reference variable box dividing value and the object value box dividing value;

The second box dividing value determining unit is used for dividing the service characteristic evaluation value, the object history evaluation value and the fixed field value into boxes according to a box dividing mode matched with the object activity of the target object to obtain corresponding service characteristic box dividing values, object history box dividing values and fixed field value box dividing values;

And the adjustable field value determining unit is used for determining the adjustable field value of the target object according to the business characteristic binning value, the object history binning value and the fixed field value binning value.

In an embodiment, when the target object is an active object, evaluating the reference variable bin value as an active reference variable bin value; the service characteristic evaluation value determining unit is specifically configured to:

In an embodiment, when the target object is an inactive object, evaluating the reference variable bin value as an inactive reference variable bin value; the service characteristic evaluation value determining unit is specifically configured to:

In an embodiment, the adjustable field value determining unit is specifically configured to:

In an embodiment, the data processing apparatus further comprises:

an adjustable field value use condition acquisition module, configured to acquire an adjustable field value use condition of the target object for an adjustable field value within an adjustable field value lifetime;

And the model adjustment module is used for dynamically adjusting the target value prediction model based on the use condition of the adjustable field value until the adjustable field value reaches the optimal value.

The data processing device provided by the embodiment of the invention can execute the data processing method provided by any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the execution method.

In one embodiment, fig. 4 is a schematic structural diagram of an electronic device implementing the data processing method according to an embodiment of the present invention, and as shown in fig. 4, a schematic structural diagram of an electronic device 10 that may be used to implement an embodiment of the present invention is shown. Electronic devices are intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. Electronic equipment may also represent various forms of mobile devices, such as personal digital processing, cellular telephones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are meant to be exemplary only, and are not meant to limit implementations of the inventions described and/or claimed herein.

As shown in fig. 4, the electronic device 10 includes at least one processor 11, and a memory, such as a Read Only Memory (ROM) 12, a Random Access Memory (RAM) 13, etc., communicatively connected to the at least one processor 11, in which the memory stores a computer program executable by the at least one processor, and the processor 11 may perform various appropriate actions and processes according to the computer program stored in the Read Only Memory (ROM) 12 or the computer program loaded from the storage unit 18 into the Random Access Memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 may also be stored. The processor 11, the ROM 12 and the RAM 13 are connected to each other via a bus 14. An input/output (I/O) interface 15 is also connected to bus 14.

Various components in the electronic device 10 are connected to the I/O interface 15, including: an input unit 16 such as a keyboard, a mouse, etc.; an output unit 17 such as various types of displays, speakers, and the like; a storage unit 18 such as a magnetic disk, an optical disk, or the like; and a communication unit 19 such as a network card, modem, wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information/data with other devices via a computer network, such as the internet, and/or various telecommunication networks.

The processor 11 may be a variety of general and/or special purpose processing components having processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), various specialized Artificial Intelligence (AI) computing chips, various processors running machine learning model algorithms, digital Signal Processors (DSPs), and any suitable processor, controller, microcontroller, etc. The processor 11 performs the various methods and processes described above, such as data processing methods.

In some embodiments, the data processing method may be implemented as a computer program tangibly embodied on a computer-readable storage medium, such as the storage unit 18. In some embodiments, part or all of the computer program may be loaded and/or installed onto the electronic device 10 via the ROM 12 and/or the communication unit 19. One or more of the steps of the data processing method described above may be performed when the computer program is loaded into RAM 13 and executed by processor 11. Alternatively, in other embodiments, the processor 11 may be configured to perform the data processing method in any other suitable way (e.g. by means of firmware).

Various implementations of the systems and techniques described here above may be implemented in digital electronic circuitry, integrated circuit systems, field Programmable Gate Arrays (FPGAs), application Specific Integrated Circuits (ASICs), application Specific Standard Products (ASSPs), systems On Chip (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and/or combinations thereof. These various embodiments may include: implemented in one or more computer programs, the one or more computer programs may be executed and/or interpreted on a programmable system including at least one programmable processor, which may be a special purpose or general-purpose programmable processor, that may receive data and instructions from, and transmit data and instructions to, a storage system, at least one input device, and at least one output device.

A computer program for carrying out methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus, such that the computer programs, when executed by the processor, cause the functions/acts specified in the flowchart and/or block diagram block or blocks to be implemented. The computer program may execute entirely on the machine, partly on the machine, as a stand-alone software package, partly on the machine and partly on a remote machine or entirely on the remote machine or server.

In the context of the present invention, a computer-readable storage medium may be a tangible medium that can contain, or store a computer program for use by or in connection with an instruction execution system, apparatus, or device. The computer readable storage medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. Alternatively, the computer readable storage medium may be a machine readable signal medium. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

To provide for interaction with a user, the systems and techniques described here can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to a user; and a keyboard and a pointing device (e.g., a mouse or a trackball) through which a user can provide input to the electronic device. Other kinds of devices may also be used to provide for interaction with a user; for example, feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form, including acoustic input, speech input, or tactile input.

The systems and techniques described here can be implemented in a computing system that includes a background component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front-end component (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such background, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: local Area Networks (LANs), wide Area Networks (WANs), blockchain networks, and the internet.

The computing system may include clients and servers. The client and server are typically remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server can be a cloud server, also called a cloud computing server or a cloud host, and is a host product in a cloud computing service system, so that the defects of high management difficulty and weak service expansibility in the traditional physical hosts and VPS service are overcome.

Embodiments of the present application also provide a computer program product comprising a computer program which, when executed by a processor, implements a data processing method as provided by any of the embodiments of the present application.

Computer program product in the implementation, the computer program code for carrying out operations of the present invention may be written in one or more programming languages, including an object oriented programming language such as Java, smalltalk, C ++ and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any kind of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or may be connected to an external computer (for example, through the Internet using an Internet service provider).

Note that the above is only a preferred embodiment of the present invention and the technical principle applied. It will be understood by those skilled in the art that the present invention is not limited to the particular embodiments described herein, but is capable of various obvious changes, rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention. Therefore, while the invention has been described in connection with the above embodiments, the invention is not limited to the embodiments, but may be embodied in many other equivalent forms without departing from the spirit or scope of the invention, which is set forth in the following claims.

Claims

1. A method of data processing, comprising:

Inputting the target historical value reference variable into a pre-established target value prediction model to obtain an object value prediction value corresponding to the target object; the target value prediction model is a prediction model enabling a value forward prediction value to reach minimum and a value reverse prediction value to reach maximum;

And determining the adjustable field value of the target object according to the evaluation reference variable bin value and the object value bin value, and the pre-acquired object history evaluation value and fixed field value.

2. The method according to claim 1, further comprising, before said data normalization of each initial historical value reference variable of the pre-acquired target object to obtain a corresponding target historical value reference variable:

and acquiring the initial historical value reference variable corresponding to the target object in the historical time window.

3. The method of claim 1, wherein the training process of the target value prediction model comprises:

Randomly extracting partial data from the original sample data to serve as a corresponding training set and a corresponding testing set;

And inputting the sample data in the training set into a pre-established original value prediction model for continuous iterative training until the obtained target value prediction model enables the value forward prediction value obtained by testing the test data in the test set to be minimum and the value reverse prediction value to be maximum.

4. The method of claim 1, wherein the target value prediction model comprises: a value forward prediction model and a value reverse prediction model; the target historical value reference variables include: a value forward reference feature variable and a value reverse reference feature variable;

Correspondingly, the step of inputting the target historical value reference variable into a pre-created target value prediction model to obtain an object value prediction value corresponding to the target object comprises the following steps:

inputting the value forward reference characteristic variable into a pre-established value forward prediction model to obtain a corresponding value forward prediction value; the value forward prediction model is a prediction model enabling a value forward prediction value to reach minimum;

inputting the value reverse reference characteristic variable into a pre-created value reverse prediction model to obtain a corresponding value reverse prediction value; the value reverse prediction model is a prediction model enabling a value reverse prediction value to reach the maximum;

In the case of including at least two of the value forward predictors, determining a corresponding value forward predictor based on a sum of each of the value forward predictors, and determining a corresponding value of the object based on a difference between the value forward predictor and the value reverse predictor;

in the case of including one of the value forward predictors, determining the corresponding value of the object based on a difference between the value forward predictor and the value reverse predictor.

5. The method of claim 1, wherein the evaluating the reference variable comprises: active and inactive reference variables; the step of acquiring the evaluation reference variable matched with the object activity of the target object from the pre-created service database comprises the following steps:

Determining the corresponding object activity according to the service activity condition of the target object in the preset duration;

determining whether the target object is an active object according to the object liveness and a preconfigured liveness threshold;

When the target object is an active object, searching an active reference variable matched with the active object from the pre-created service database;

And searching an inactive reference variable matched with the inactive object from the pre-created service database when the target object is the inactive object.

6. The method of claim 1, wherein said determining an estimated reference variable binning value and an object value binning value for each of the estimated reference variables and the object value predictors comprises:

determining a corresponding target box division mode according to the variable types of the evaluation reference variable and the object value predicted value;

And carrying out box division on each evaluation reference variable and the object value predicted value by adopting the target box division mode to obtain corresponding evaluation reference variable box division values and object value box division values.

7. The method of claim 1, wherein said determining the adjustable field value of the target object based on the evaluation reference variable binning value and the object value binning value, and a pre-acquired object history evaluation value and fixed field value, comprises:

Determining a corresponding service characteristic evaluation value according to the evaluation reference variable box division value and the object value box division value;

The business characteristic evaluation value, the object history evaluation value and the fixed field value are classified according to a classification mode matched with the object activity of the target object, and corresponding business characteristic classification value, object history classification value and fixed field value classification value are obtained;

and determining the adjustable field value of the target object according to the business characteristic box value, the object history box value and the fixed field value box value.

8. The method of claim 7, wherein the evaluation reference variable binning value is an active reference variable binning value when the target object is an active object; the determining a corresponding service feature evaluation value according to the evaluation reference variable binning value and the object value binning value comprises:

And carrying out weighted summation on the service potential matching value and the first service balance matching value to obtain the corresponding service characteristic evaluation value.

9. The method of claim 7, wherein when the target object is an inactive object, the evaluation reference variable binning value is an inactive reference variable binning value; the determining a corresponding service feature evaluation value according to the evaluation reference variable binning value and the object value binning value comprises:

And carrying out weighted summation on the active rate matching value and the second service balance matching value to obtain the corresponding service characteristic evaluation value.

10. The method of claim 7, wherein said determining the adjustable field value of the target object based on the business feature binning value, the object history binning value and the fixed field value binning value comprises:

and determining an adjustable field value of the target object according to the adjustable basic value and the adjustment coefficient.

11. The method according to any one of claims 1-10, further comprising:

Acquiring an adjustable field value use condition of the target object aiming at the adjustable field value in the adjustable field value use period;

Dynamically adjusting the target value prediction model based on the adjustable field value usage until the adjustable field value is optimal.

12. A data processing apparatus, comprising:

the variable acquisition module is used for acquiring an evaluation reference variable matched with the object activity of the target object from a pre-created service database;

And the adjustable field value determining module is used for determining the adjustable field value of the target object according to the evaluation reference variable box value and the object value box value, and the pre-acquired object history evaluation value and fixed field value.

13. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable by the processor, the processor implementing the data processing method according to any one of claims 1-11 when executing the computer program.

14. A computer-readable storage medium, on which a computer program is stored, characterized in that the program, when being executed by a processor, implements a data processing method according to any one of claims 1-11.

15. A computer program product comprising a computer program, characterized in that the computer program, when executed by a processor, implements the data processing method according to any of claims 1-11.