Charging policy generation method and device, equipment, storage medium and product

Abstract

The invention discloses a charging strategy generation method and device, equipment, a storage medium and a product, and relates to the technical field of infrastructure and IT support, and the charging strategy generation method comprises the steps: obtaining target user ticket data collected in real time, and user elements and charging dimension data stored in a cache database; inputting the target user ticket data, the user elements and the charging dimension data into an updated preset charging strategy generation model for feature extraction to obtain a user data vector and a user demand vector; and determining an attention weight based on the user data vector and the preset charging policy generation model, and generating a charging policy based on the attention weight, the user data vector and the user demand vector to obtain a target charging policy. According to the method and the device, the problem that the change of the on-demand use condition of the user cannot be responded in real time due to the fact that the related technologies mostly depend on simple rule setting or limited data analysis can be avoided.

Description

Technical Field

[0001] The present application relates to the technical field of infrastructure and IT support, and in particular to a billing strategy generation method, device, equipment, storage medium and product. Background Art

[0002] With the rapid development of information technology, communication services have become an indispensable part of people's daily lives. Traditional billing methods are usually based on fixed price models, such as charging by minutes and traffic usage. However, this static billing model is difficult to meet the increasingly diverse and personalized needs of users, especially in the context of the mobile Internet era, where users' consumption habits and service needs change rapidly. Traditional billing methods are too rigid and cannot effectively reflect the real value contribution of users. At the same time, they also limit the ability of operators to provide more personalized services.

[0003] In the related technology, some improved billing methods have begun to try to combine the actual usage of users, such as adjusting the charging standards by analyzing the user's behavior patterns. However, most of these methods rely on simple rule settings or limited data analysis, and cannot respond to changes in user on-demand usage in real time. The corresponding billing rules are not applicable to the unique needs of different users and lack sufficient flexibility. Summary of the invention

[0004] The main purpose of this application is to provide a billing strategy generation method, which aims to solve the problem of adjusting charging standards by analyzing user behavior patterns. However, most of these methods rely on simple rule settings or limited data analysis, and cannot respond to changes in user on-demand usage in real time. The corresponding billing rules are also not suitable for the unique needs of different users, and lack sufficient flexibility.

[0005] To achieve the above purpose, the present application proposes a charging strategy generation method, the charging strategy generation method comprising:

[0006] Obtain the target user's call list data collected in real time and the user elements and billing dimension data stored in the cache database;

[0007] Inputting the target user call list data, the user elements and the billing dimension data into the updated preset billing strategy generation model for feature extraction to obtain a user data vector and a user demand vector;

[0008] Based on the user data vector and the preset billing strategy generation model, an attention weight is determined, and a billing strategy is generated based on the attention weight, the user data vector and the user demand vector to obtain a target billing strategy.

[0009] In one embodiment, the step of determining the attention weight based on the user data vector and the preset charging strategy generation model, generating the charging strategy based on the attention weight, the user data vector and the user demand vector, and obtaining the target charging strategy includes:

[0010] Based on the preset parameters of the attention mechanism layer of the preset billing strategy generation model and the user data vector, generate an attention score for each vector in the user data vector, wherein the preset parameters are obtained based on the initial parameter optimization;

[0011] Normalizing the attention score to obtain an attention weight;

[0012] Combining the attention weight, the user data vector, and the user demand feature vector through addition pooling to obtain a pooled first data vector;

[0013] Performing user consumption behavior learning on the first data vector to obtain a second data vector including an updated user consumption pattern;

[0014] Interacting all features in the second data vector to obtain an interacted third data vector;

[0015] A target strategy corresponding to the third data vector is generated through a preset mapping relationship between the preset strategy and the data vector.

[0016] In one embodiment, the billing dimension data includes a first personalized dimension or a second personalized dimension, the user element includes a first user element or a second user element, and the step of acquiring the target user call list data collected in real time and the user element and billing dimension data stored in the cache database includes:

[0017] If the user terminal does not initiate a user demand, obtain the user usage record within a preset time, extract information through the historical user call list data in the user usage record, obtain the first user factor, and determine the first personalized dimension obtained based on the analysis of the user usage record;

[0018] If the user terminal initiates the user demand, a second personalized dimension is determined based on the user demand, and the user usage record within a preset time is obtained, and information is extracted through historical user call list data in the user usage record to obtain a second user element;

[0019] The user elements and the charging dimension data are stored in a cache database.

[0020] In one embodiment, the step of storing the user element and the billing dimension data in a cache database includes:

[0021] Calculate the mean and standard deviation of the numerical data in the historical user call list data, the user elements and the billing dimension data;

[0022] Based on the mean and the standard deviation, the numerical data is scaled to a preset range to obtain the scaled historical user call list data, the user elements and the billing dimension data;

[0023] Obtaining a charging history strategy corresponding to the historical user call list data, and using the historical user call list data, the user element and the charging dimension data as sample data, wherein the historical strategy is a sample label corresponding to the sample data;

[0024] Obtaining an initial billing policy generation model;

[0025] The sample data and the sample labels are input into the initial charging policy generation model for iterative training to obtain a preset charging policy generation model that meets the accuracy requirements.

[0026] In one embodiment, the step of inputting the sample data and the sample label into the initial charging policy generation model for iterative training to obtain a preset charging policy generation model that meets the accuracy requirement includes:

[0027] Based on the input layer of the initial charging policy generation model, converting the sample data into multi-dimensional user-related data, and outputting the user-related data to the embedding layer of the initial charging policy generation model;

[0028] Based on the embedding layer, the user data and user needs in the user-related data are parsed, the parsed user-related data is converted into a preset base vector, and the preset base vector corresponding to the user data and the preset base vector corresponding to the user need are spliced ​​respectively, and the spliced ​​user data vector and user need vector are input into the attention mechanism layer of the initial billing strategy generation model;

[0029] Based on the attention mechanism layer and the initial parameters of the attention mechanism layer, the user data vector and the user demand feature vector are combined by addition pooling, and the weighted first sample vector is input into the memory network layer of the initial charging strategy generation model;

[0030] Based on the memory network layer and the initial weight matrix and initial bias vector of the memory network layer, the first sample vector is subjected to user consumption behavior learning, and the learned second sample vector including the updated user consumption pattern is input into the interaction layer of the initial charging strategy generation model;

[0031] Based on the interaction layer, all features in the second sample vector are interacted, and the third sample vector after interaction is input into the decision output layer of the initial charging strategy generation model;

[0032] Based on the decision output layer, performing strategy mapping on the third sample vector to obtain an initial charging strategy;

[0033] Based on the sample labels and the initial charging strategy, the initial parameters, initial weight matrix and initial bias vector are optimized to obtain a preset charging strategy generation model that meets the accuracy requirements after optimization.

[0034] In one embodiment, the step of inputting the sample data and the sample label into the initial charging policy generation model for iterative training to obtain a preset charging policy generation model that meets the accuracy requirement includes:

[0035] Storing the preset charging policy generation model in a shared memory library;

[0036] Based on a preset timing time, the preset charging policy generation model stored in the shared memory library is updated to obtain an updated preset charging policy generation model.

[0037] In addition, to achieve the above purpose, the present application also proposes a charging policy generation device, the charging policy generation device comprising:

[0038] An acquisition module is used to acquire the target user's call list data collected in real time and the user elements and billing dimension data stored in the cache database;

[0039] An extraction module, used for inputting the target user call list data, the user elements and the billing dimension data into the updated preset billing strategy generation model for feature extraction to obtain a user data vector and a user demand vector;

[0040] A generation module is used to determine the attention weight based on the user data vector and the preset billing strategy generation model, and generate a billing strategy based on the attention weight, the user data vector and the user demand vector to obtain a target billing strategy.

[0041] In addition, to achieve the above-mentioned purpose, the present application also proposes a billing policy generation device, which includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the billing policy generation method described above.

[0042] In addition, to achieve the above objectives, the present application also proposes a storage medium, which is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, the steps of the billing policy generation method described above are implemented.

[0043] In addition, to achieve the above-mentioned purpose, the present application also provides a computer program product, which includes a computer program. When the computer program is executed by a processor, the steps of the charging policy generation method described above are implemented.

[0044] One or more technical solutions proposed in this application have at least the following technical effects:

[0045] Compared with the related art, the charging standards are adjusted by analyzing the user's behavior patterns. However, most of these methods rely on simple rule settings or limited data analysis, and cannot respond to changes in user on-demand usage in real time. The corresponding billing rules are not applicable to the unique needs of different users and lack sufficient flexibility. In this application, the target user call list data collected in real time and the user elements and billing dimension data stored in the cache database are obtained; the target user call list data, the user elements and the billing dimension data are extracted through the updated preset billing strategy generation model to obtain user data vectors and user demand vectors based on the user Data vector, determine the attention weight, generate a billing strategy based on the attention weight, the user data vector and the user demand vector, and obtain a target billing strategy. It can be understood that the present application uses an updated preset billing strategy generation model to extract features from the target user's call list data, user elements and billing dimension data to obtain a user data vector and a user demand vector, and introduces the user demand vector into the attention mechanism to generate a billing strategy that matches the user data vector and the user demand vector. This can avoid the problem that most related technologies rely on simple rule settings or limited data analysis and cannot respond to changes in user on-demand usage in real time. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0047] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0048] Figure 1 A flowchart of the first embodiment of the method for generating a charging strategy for this application is provided;

[0049] Figure 2 This is a billing policy matching flow chart of the billing policy generation method for this application;

[0050] Figure 3 A calculation power charging formula rule diagram for the billing strategy generation method of this application;

[0051] Figure 4 A flowchart of the second embodiment of the method for generating a charging strategy for this application is provided;

[0052] Figure 5 A model structure diagram of the billing strategy generation method for this application;

[0053] Figure 6 This is a schematic diagram of the module structure of the charging strategy generating device according to an embodiment of the present application;

[0054] Figure 7 A schematic diagram of the device structure of the hardware operating environment involved in the billing policy generation method in the embodiment of the present application.

[0055] The purpose, features and advantages of this application will be further described in conjunction with the embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0056] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and are not used to limit the present application.

[0057] In order to better understand the technical solution of the present application, a detailed description will be given below in conjunction with the accompanying drawings and specific implementation methods.

[0058] The main solution of the embodiment of the present application is: obtaining the target user call record data collected in real time and the user elements and billing dimension data stored in the cache database; inputting the target user call record data, the user elements and the billing dimension data into the updated preset billing policy generation model for feature extraction to obtain the user data vector and the user demand vector; determining the attention weight based on the user data vector and the preset billing policy generation model, generating the billing strategy based on the attention weight, the user data vector and the user demand vector to obtain the target billing strategy.

[0059] In the related technology, some improved billing methods have begun to try to combine the actual usage of users, such as adjusting the charging standards by analyzing the user's behavior patterns. However, most of these methods rely on simple rule settings or limited data analysis, and cannot respond to changes in user on-demand usage in real time. The corresponding billing rules are not applicable to the unique needs of different users and lack sufficient flexibility.

[0060] The present application uses an updated preset billing strategy generation model to perform feature extraction on the target user's call record data, user factors and billing dimension data to obtain a user data vector and a user demand vector, and introduces the user demand vector into the attention mechanism to generate a billing strategy that matches the user data vector and the user demand vector. This can avoid the problem that most related technologies rely on simple rule settings or limited data analysis and cannot respond to changes in user on-demand usage in real time.

[0061] It should be noted that the execution subject of this embodiment can be a computing service device with data processing, network communication and program running functions, such as a tablet computer, a personal computer, a mobile phone, etc., or an electronic device capable of realizing the above functions, etc. The following takes a charging policy generation device as an example to illustrate this embodiment and the following embodiments.

[0062] Based on this, the present application embodiment provides a method for generating a charging strategy, referring to Figure 1 , Figure 1 This is a flow chart of the first embodiment of the method for generating a charging policy for this application.

[0063] In this embodiment, refer to Figure 2 , Figure 2 A charging policy matching flow chart is provided, and the charging policy generation method includes steps S100 to S300:

[0064] Step S100, acquiring the target user's call list data collected in real time and the user elements and billing dimension data stored in the cache database;

[0065] It should be noted that the real-time subject of this embodiment is a billing policy generation device. The billing policy generation device is provided with a cache database, which is used to store user factors and billing dimension data. The target user call list includes data such as user geographic location, service type and usage. User factors include factors such as user type, user credit rating and historical consumption level. Billing dimension data includes pricing dimension, control dimension, personalized dimension and billing dimension, wherein each dimension in the billing dimension data can be represented by a corresponding code.

[0066] It is understandable that the billing strategy generation device obtains the real-time collected target user call record data based on the API interface, and obtains user elements and billing dimension data from the cache database. Among them, the API interface mainly provides real-time prediction capabilities for new orders or special requirements of the day to collect user call records in real time.

[0067] Step S200, inputting the target user call list data, the user elements and the charging dimension data into the updated preset charging strategy generation model for feature extraction to obtain a user data vector and a user demand vector;

[0068] It should be noted that the user demand vector is determined based on the personalized dimension. The preset billing policy generation model is a model that is updated periodically. The billing policy generation device extracts features from the target user's call list data, user elements, and billing dimension data through the One-HotEncoding method in the preset billing policy generation model to obtain the user data vector and the user demand vector determined based on the personalized dimension.

[0069] Specifically, the model input data format information is as follows:

[0070] {service_usage:["BeiJing",111712,100],user_profile:[0,601,0.5],dimension_features:[6000008,6000008002]};

[0071] The meanings of the different dimensions in the above example are:

[0072] {service_usage:[user location, service type, usage], user_profile:[user type, user credit rating, historical consumption level], dimension_features:[billing dimension code, billing pricing code]}.

[0073] Furthermore, the charging policy generation device converts the user's geographic location "Beijing" into an integer encoding of 110 through the One-Hot Encoding method, and outputs other features as they are. The converted data format is as follows:

[0074] [110,111712,100,0,601,0.5,6000008,6000008002,[110,111712,100,0.5]].

[0075] Step S300, based on the user data vector and the preset billing strategy generation model, determine the attention weight, generate the billing strategy based on the attention weight, the user data vector and the user demand vector, and obtain the target billing strategy.

[0076] It is understandable that the target charging strategy is an optimized charging strategy obtained based on the real-time target user call record. The charging strategy generation device determines the attention weight of the model based on the user data vector, and generates the optimized target charging strategy based on the real-time target user call record based on the obtained attention weight, user data vector and user demand vector.

[0077] In a feasible implementation manner, based on the user data vector and the preset charging strategy generation model, an attention weight is determined, and a charging strategy is generated based on the attention weight, the user data vector, and the user demand vector, and the step of obtaining a target charging strategy includes the following steps:

[0078] Based on the preset parameters of the attention mechanism layer of the preset billing strategy generation model and the user data vector, generate an attention score for each vector in the user data vector, wherein the preset parameters are obtained based on the initial parameter optimization;

[0079] It should be noted that the billing policy generation device calculates the attention score of each feature vector in the user data vector through the preset parameters of the attention mechanism layer of the preset billing policy generation model.

[0080] Specifically, the charging policy generation device calculates and obtains each feature vector h in the user data vector H i The attention score of h i is the i-th feature representation in the user data vector H, and outputs e ti is the attention score of the feature vector of the user data vector H, let s t is the parameter of the attention mechanism layer, which is the parameter obtained by model training, and outputs e ti is the attention score of the feature vector of the user data vector H. The calculation formula is as follows:

[0081]

[0082] Normalizing the attention score to obtain an attention weight;

[0083] It is understandable that the attention weight is the probability distribution of the attention score. The charging policy generation device normalizes the attention score of the user data vector, calculates its probability distribution, and assigns a corresponding attention weight to each feature based on the probability distribution.

[0084] Specifically, the charging strategy generates the device’s attention score for the output ti Use the softmax function to normalize and calculate its probability distribution. Give each feature a corresponding weight based on the probability distribution, with a value between 0 and 1, and output the attention weight vector a t , where the basic principle of the softmax function is to convert a set of values ​​into a probability distribution, each value is converted into a relative probability, and the sum of these probabilities is equal to 1. The formula is as follows:

[0085] a t =softmax(e t )

[0086] Combining the attention weight, the user data vector, and the user demand feature vector through addition pooling to obtain a pooled first data vector;

[0087] It should be noted that the charging strategy generation device combines the attention weight, the user data vector and the user demand feature vector through additive pooling to obtain a pooled first data vector.

[0088] Specifically, the charging strategy generation device adds the attention weight vector a through addition pooling. t Combined with the user demand feature vector q, we can get a vector that changes dynamically based on user demand. The output vector c after pooling is t , and its calculation formula is as follows:

[0089]

[0090] Among them, c t is the output vector after pooling; h i is the i-th feature representation in the user data vector H; b1, b2 are the bias vectors of this layer, which are optimized and updated by the back-propagation gradient descent method and are used to compare with the features in the input data.

[0091] Performing user consumption behavior learning on the first data vector to obtain a second data vector including an updated user consumption pattern;

[0092] It is understandable that the second data vector is a new user data vector that takes into account user needs and includes an updated user consumption pattern. The charging policy generation device performs user consumption behavior learning on the first data vector to obtain a new user data vector that includes an updated user consumption pattern after learning.

[0093] Interacting all features in the second data vector to obtain an interacted third data vector;

[0094] It should be noted that the charging policy generating device performs a second-order interaction between all features of the new user data vector, that is, calculates the element-by-element product of the user data features to obtain a third data vector after interaction.

[0095] A target strategy corresponding to the third data vector is generated through a preset mapping relationship between the preset strategy and the data vector.

[0096] It is understandable that the preset mapping relationship is pre-stored and used to characterize the mapping relationship between the policy and the vector. The target policy is the policy corresponding to the third data vector. The charging policy generation device maps and searches the policy corresponding to the third data vector through the preset mapping relationship to obtain the target policy corresponding to the third data vector.

[0097] In particular, after obtaining the target strategy, the billing strategy generation device can further predict future call bills under different strategies based on various strategies and historical call bills. Since the billing strategy output by the model is a calculation power calculation formula rule, it is expressed as a billing formula. According to the formula, the user's current consumption amount can be calculated. Figure 3 , Figure 3 A diagram of the calculation power and fee formula is provided.

[0098] In this embodiment, the billing strategy generation device generates user data vectors and user demand feature vectors based on the constructed intelligent dynamic billing model, through user call records, user factors and billing dimension data, using the One-Hot Encoding method, and additionally introduces user demand feature vectors to optimize the attention mechanism, thereby generating a target billing strategy that matches the user data and user needs.

[0099] The present application embodiment provides a method for generating a charging strategy. Figure 4 , Figure 4 This is a flow chart of the second embodiment of the method for generating a charging policy for this application.

[0100] In a feasible implementation manner, the step of obtaining the target user call list data collected in real time and the user elements and billing dimension data stored in the cache database includes steps A100 to A300:

[0101] Step A100: if the user terminal does not initiate a user demand, obtain a user usage record within a preset time, extract information through historical user call list data in the user usage record, obtain user factors, and determine a first personalized dimension obtained based on the analysis of the user usage record;

[0102] It should be noted that the first personalized dimension is the user demand determined by the billing policy generation device, which can be various possible user demands by default in the billing policy generation device, or can accept the user demand on the user side. If the client corresponding to the user side does not initiate a request for personalized services based on the user demand, the billing policy generation device will automatically identify the user's consumption pattern and preference, and determine which type of billing method the user is more likely to accept based on this, and use this type of billing method as the first personalized dimension, and extract information through historical user call data to obtain user elements. Due to the special nature of user elements, the user elements of the same user remain basically unchanged.

[0103] Among them, user needs include the following typical needs:

[0104] a) Fixed payment demands represented by annual and monthly packages;

[0105] b) There is no prepaid fixed fee, and the consumption is based on the actual consumption of calls, traffic and other services or resources, and the demand for pay-as-you-go;

[0106] c) Setting different charging standards for different peak and valley periods according to the peak and valley payment needs;

[0107] d) Personalized customized service function modules require customized payment only for the service function modules used.

[0108] It can be understood that the above four user demands a) to d) are only used as examples. In the process of providing services, the billing policy generation device can provide various new demands, collect demands from the user side and finally launch products and services that meet different demands.

[0109] In particular, when a user signs a contract with the service provider corresponding to the billing policy generating device, the call records that have occurred will inevitably point to one of the four typical user needs a) to d) mentioned above, and the user needs in the personalized dimension can be determined based on the call records that occurred after the user signed the contract with the service provider (for example, one of the four typical user needs a) to d) mentioned above).

[0110] Step A200: if the user terminal initiates the user demand, a second personalized dimension is determined based on the user demand, and the user usage record within a preset time is obtained, and information is extracted through historical user call list data in the user usage record to obtain user factors;

[0111] It is understandable that the second personalized dimension is determined based on the user demand initiated by the user terminal corresponding to the user. If the charging policy generating device receives a request for personalized service initiated by the user terminal for user demand, it stores it in the data object indicated by the personalized dimension, wherein the personalized dimension and the remaining dimension are both encoded.

[0112] In particular, in order to avoid frequent modification of existing codes, each code can point to a specific data object. Since the personalized dimension indicates the user demand initiated by the user side, assuming that the user demand feature vector initiated by the user side is no longer [1,0,0,0] but [0,1,0,0] or even [0,1,1,1], then based on the user demand initiated by the user and the user call records that have occurred, the intelligent billing strategy output by the model can provide users with new personalized service options. The specific choice is up to the user.

[0113] Specifically, the billing strategy generation device can accurately identify users who pay a fixed monthly or annual fee, and apply the corresponding billing strategy, for example, a monthly / annual fixed fee dimensional billing method to meet the user's demand for stable costs. The following specific examples of specific implementation steps for a variety of typical user needs, but this application is not limited to the following user needs:

[0114] User requirement 1: Fixed budget, expected stable cost

[0115] The billing strategy generation device determines whether a user pays a fixed monthly fee or an annual fee mainly by analyzing the user's historical consumption data and behavior patterns. This is achieved through the following steps:

[0116] a) Data collection and preprocessing: Collect user usage records over the past period of time (the last three months), including but not limited to call data, service type, and usage volume, and combine user factor data such as user type, credit rating, and historical consumption level information;

[0117] b) Setting judgment thresholds: Based on the above characteristics, judgment thresholds are set. Consecutive consumption records of the same or similar amounts in the same month or annual cycle can be used as indicators of paying fixed monthly or annual fees.

[0118] The billing policy generation device calculates the amount of consumption of the user within a certain time window and uses statistical methods (standard deviation) to evaluate the fluctuation degree of the consumption amount. If the fluctuation of the consumption amount is small, it indicates that the consumption amount is relatively stable.

[0119] c) Model training and prediction: Using the intelligent dynamic billing model, the model is trained based on the labeled user consumption behavior data set. After training, the model can predict whether the user will pay a fixed monthly fee or an annual fee based on the characteristics of the user's consumption record;

[0120] Furthermore, through the above steps, the billing strategy generation device can accurately identify users who pay fixed monthly or annual fees, and thus apply corresponding billing strategies, such as monthly / annual fixed fee-based billing methods, to meet users' demand for stable costs.

[0121] User requirement 2: Pay by usage and flexibly control costs

[0122] The model obtains a pay-as-you-go billing strategy based on user needs. Users only pay for the services or resources they actually consume, without prepaying a fixed fee. Users only pay for the services or resources they actually consume. The judgment is made through the following methods:

[0123] a) Extract key information from the user's historical consumption data for the past three months, including but not limited to service type, usage, and time. These data usually come from the user's call records and reflect the user's actual consumption of various services at different time points;

[0124] b) Analyze the consumption patterns of users, including consumption frequency, usage fluctuations, and consumption time periods; for users who pay by usage, their consumption records often show irregular and irregular consumption behaviors. Each consumption record is directly related to the actual amount of services or resources used, and there are no fixed periodic large payment records; specifically, the consumption frequency is measured by calculating the number of transactions made by users within a certain time window, and the degree of usage fluctuation is evaluated through statistical methods (standard deviation).

[0125] c) Based on the user's consumption pattern, set indicators to determine whether the user pays according to usage. If the user's consumption record shows small, frequent transactions without obvious periodic large expenditures, it is highly consistent with the characteristics of pay-as-you-go;

[0126] Set a threshold for small transactions, such as spending less than 50 or 100 yuan;

[0127] Define the threshold for frequent transactions, such as more than 10 transactions in a month;

[0128] Check for recurring large expenses, such as large purchases on fixed dates each month.

[0129] d) Using a machine learning model, a model trained based on historical consumption data is used to predict whether the user is inclined to pay based on usage; the model input includes the user's past usage behavior and consumption characteristics, and the output is the prediction result of the user's payment model.

[0130] Step A300: storing the user elements and the billing dimension data in a cache database.

[0131] It should be noted that after obtaining the user factor and the billing dimension, the charging policy generating device will store the above data in the cache database for subsequent calling.

[0132] In a feasible implementation manner, after the step of storing the user element and the billing dimension data in a cache database, the step includes:

[0133] Calculate the mean and standard deviation of the numerical data in the historical user call list data, the user elements and the billing dimension data;

[0134] It is understandable that the charging policy generating device needs to calculate the mean and standard deviation of each user feature in the entire user historical usage data information when performing normalization and scaling.

[0135] Specifically, there is a feature matrix k, in which each row represents a sample and each column represents a feature. For each feature, the charging policy generating device calculates its mean (mean) and standard deviation (standard_deviation) in the entire data set.

[0136] Based on the mean and the standard deviation, the numerical data is scaled to a preset range to obtain the scaled historical user call list data, the user elements and the billing dimension data;

[0137] It should be noted that the preset range is [0, 1]. For numerical data, the charging policy generating device needs to determine to scale the data to [0, 1] in order to achieve standardized scaling of the data, so as to scale all numerical feature values ​​to between 0 and 1.

[0138] Specifically, the charging policy generation device applies a standardized scaling formula to the numerical data to scale all values ​​to within the target range. The standardized scaling formula is as follows:

[0139]

[0140] Among them, k is the value of the original feature, mean(k) is the mean of the feature set, σ(k) is the standard deviation of the feature set, and m is the converted value.

[0141] Obtaining a charging history strategy corresponding to the historical user call list data, and using the historical user call list data, the user element and the charging dimension data as sample data, wherein the historical strategy is a sample label corresponding to the sample data;

[0142] It is understandable that the sample data is the user call data, user elements and billing dimension data of the last three months. The billing strategy generation device obtains the user call data, user elements and billing dimension data of the last three months as sample data, and obtains the billing history policy corresponding to the historical user call data as a sample label.

[0143] Obtaining an initial billing policy generation model;

[0144] It should be noted that the initial charging policy generation model is an intelligent dynamic charging model (Intelligent Dynamic Fee Management Model, IDFM). The charging policy generation device obtains the intelligent dynamic charging model, referring to Figure 5 , Figure 5 A diagram of the model structure is provided.

[0145] The sample data and the sample labels are input into the initial charging policy generation model for iterative training to obtain a preset charging policy generation model that meets the accuracy requirements.

[0146] It is understandable that the initial charging policy generation model is trained based on sample data with labeled sample labels to obtain a preset charging policy generation model that meets the accuracy requirements.

[0147] Specifically, the billing policy generation device divides the sample data into a training set, a validation set, and a test set. 70% of the data is used for training, 15% of the data is used for validation, and 15% of the data is used for testing. The difference between the test set and the validation set is that they have different purposes: the validation set is mainly used to adjust model parameters and select models, help determine the complexity of the model, and prevent overfitting. The test set is used to finally evaluate the performance of the model, with the purpose of evaluating the generalization ability of the model on unknown data.

[0148] In a feasible implementation manner, the step of inputting the sample data and the sample label into the initial charging policy generation model for iterative training to obtain a preset charging policy generation model that meets the accuracy requirements includes the following steps:

[0149] Based on the input layer of the initial charging policy generation model, converting the sample data into multi-dimensional user-related data, and outputting the user-related data to the embedding layer of the initial charging policy generation model;

[0150] It should be noted that user-related data includes basic information and behavior data of users, and information related to billing includes consumption amount, consumption frequency, user credit rating and user geographic location. The billing policy generation device receives user call records, user elements and billing dimensions through the input layer of the initial billing policy generation model and outputs them as multi-dimensional user-related data, wherein the user-related data includes basic information and behavior data of users, and information related to billing includes consumption amount, consumption frequency, user credit rating and user geographic location.

[0151] Based on the embedding layer, the user data and user needs in the user-related data are parsed, the parsed user-related data is converted into a preset base vector, and the preset base vector corresponding to the user data and the preset base vector corresponding to the user need are spliced ​​respectively, and the spliced ​​user data vector and user need vector are input into the attention mechanism layer of the initial billing strategy generation model;

[0152] It is understandable that the preset base is binary. The charging strategy generation device parses the user data and user needs in the multi-dimensional user-related data output by the input layer through the embedding layer of the initial charging strategy generation model, and then uses the One-Hot Encoding method to perform feature splicing according to the parsed multi-dimensional user-related data, and splices out the user data vector and the user demand feature vector. The charging strategy generation device is not only based on the user call bill that has occurred, but also based on the user demand, so as to provide personalized services and flexible and accurate billing and more predictive billing strategies. Therefore, the input layer receives inputs such as user call bills, user elements and billing dimensions, and the personalized dimensions included in the billing dimensions indicate user needs, and then, the user data vector and the user demand feature vector are spliced ​​through the embedding layer.

[0153] Based on the attention mechanism layer and the initial parameters of the attention mechanism layer, the user data vector and the user demand feature vector are combined by addition pooling, and the weighted first sample vector is input into the memory network layer of the initial charging strategy generation model;

[0154] It should be noted that the charging policy generating device combines the user data vector and the user demand feature vector by adding pooling according to the initial parameters to obtain a weighted first sample vector.

[0155] Based on the memory network layer and the initial weight matrix and initial bias vector of the memory network layer, the first sample vector is subjected to user consumption behavior learning, and the learned second sample vector including the updated user consumption pattern is input into the interaction layer of the initial charging strategy generation model;

[0156] It is understandable that the memory network layer is an LSTM layer (Long Short-Term Memory Layer). The billing policy generation device gradually learns the user consumption behavior pattern through the LSTM layer of the initial billing policy generation model, the weighted user data feature vector output by the attention mechanism layer, through the forget gate calculation, input gate calculation, current layer parameter update and output gate calculation in the long short-term memory neural network. The pattern is not only used to predict future consumption behavior, but also used to generate: a new user data vector x that takes into account user needs and contains updated user consumption patterns, where: x t-1 represents the updated user data vector at time t-1, W represents the weight matrix of this layer, b represents the bias vector of this layer, which are all optimized and updated by the back-propagation gradient descent method, C t-1 represents the context vector at time t-1, which means storing long-term dependency information, h t Represents the input user data feature vector at time t. At time t, the corresponding calculation formula is as follows:

[0157] x,C t =LSTM(x t-1 ,C t-1 ,h t ; W, b)

[0158] Based on the interaction layer, all features in the second sample vector are interacted, and the third sample vector after interaction is input into the decision output layer of the initial charging strategy generation model;

[0159] It should be noted that the charging policy generation device performs second-order interactions between all features on the new user data vector x output by the LSTM layer through the interaction layer of the initial charging policy generation model, that is, calculates the element-by-element product of the user data features. The output of this layer is a user data vector z containing second-order cross terms between features. The specific formula is as follows:

[0160]

[0161] Among them, let v be the weight factor of this layer, w i is the weight matrix of the i-th feature, w j is the weight matrix of the jth feature, and its initial value is set to 0.9. y is the second-order cross of all features of the user data, and ⊙ represents element multiplication. is the activation function, let w a The weight matrix of this layer. Parameters v, w i and w aThe update is optimized by the back-propagation gradient descent method. Suppose the user data vector x is input to the interaction layer (it can be understood that the user data vector at this time has passed through the previous attention mechanism layer and LSTM layer. x contains n features, then the input dimension of the interaction layer is n, and the output dimension of the interaction layer is n+C(n,2)=n+n(n-1) / 2.

[0162] If the input x vector is:

[0163] x=[110,111712,100,0,601,0.5,6000008,6000008002,[110,111712,100,0.5]];

[0164] The output vector z is as follows:

[0165] z=[110,111712,100,0,601,0.5,6000008,6000008002,x 1 ·x 2 ,x 1 ·x 3 ,...,x 7 ·

[0166] x 8 ].

[0167] Based on the decision output layer, performing strategy mapping on the third sample vector to obtain an initial charging strategy;

[0168] It can be understood that the charging strategy generation device performs feature mapping integration processing on the user data vector z containing the second-order cross terms between features output by the interaction layer through a fully connected layer (Fully Connected Layer) plus a softmax activation function through the decision output layer (Decision Layer) of the initial charging strategy generation model, and outputs the final model prediction result. The charging strategy output by the decision output layer is f, and the corresponding formula of f is as follows:

[0169] f = softmax(Vz+b′)

[0170] Among them, V,b′ is the parameter factor of this layer, and its value range is [0,1]. The initial value of both is set to 0.5, and the back-propagation gradient descent method is used to optimize the update.

[0171] Based on the sample labels and the initial charging strategy, the initial parameters, initial weight matrix and initial bias vector are optimized to obtain a preset charging strategy generation model that meets the accuracy requirements after optimization.

[0172] It should be noted that the loss function of the initial billing policy generation model of the billing policy generation device uses the mean square error (MSE) function and the Adam optimizer, sets 0.01 as the starting learning rate, uses the adaptive learning rate optimizer to automatically adjust the learning rate of each parameter according to the historical gradient information, sets the batch value to 20 according to the data set (2,000,000 data items), and randomly divides the data set into 7:1.5:1.5 batches, i.e., 1,400,000 training sets, 300,000 test sets, and validation sets, with 10,000 iterations. In each iteration, the training set is used for forward propagation to calculate the predicted value, and the model parameters are updated through back propagation. At the same time, the validation set is used to monitor the model performance to avoid overfitting. Finally, the test data set is used to give an independent evaluation of the expected performance of the model in actual applications, and finally an effective and reliable trained intelligent dynamic billing model is obtained.

[0173] In a feasible implementation manner, after the step of inputting the sample data and the sample label into the initial charging policy generation model for iterative training to obtain a preset charging policy generation model that meets the accuracy requirement, the step includes:

[0174] Storing the preset charging policy generation model in a shared memory library;

[0175] It is understandable that the shared memory library can improve the access speed and efficiency of the model, thereby quickly predicting and adjusting user needs and billing dimensions. The billing policy generation device loads the generated model into the shared memory library.

[0176] Based on a preset timing time, the preset charging policy generation model stored in the shared memory library is updated to obtain an updated preset charging policy generation model.

[0177] It should be noted that since new users who place orders or have special requirements that require retraining every day may not be included in the prediction results or the prediction results may no longer be the latest, the matching may not be the optimal billing strategy. Therefore, the billing policy generation device will be based on a preset timer (for example, 0:30) and will be trained and updated to the latest billing policy generation model in the system every day.

[0178] In particular, the billing policy generation model will encapsulate the trained model into a Docker image, deploy it as a microservice using Kubernetes or similar container orchestration tools, implement high-performance API services, and reduce the model size and improve the inference speed through model quantization, pruning and other technologies. At the same time, high-frequency requests or results are cached to reduce unnecessary repeated calculations.

[0179] Specifically, the specific indicators of the model's real-time prediction mainly focus on the model's performance, response speed, and stability. The following are several key real-time prediction indicators:

[0180] Response Time: The time it takes from the model receiving input data to generating predicted output. In real-time prediction scenarios, low response time is crucial because it directly affects user experience and system performance.

[0181] Throughput: describes the number of requests that the model can handle per unit time. For highly concurrent real-time systems, high throughput means that the model can efficiently handle a large number of requests that arrive at the same time.

[0182] Model Stability: The consistency and predictability of the model output, especially in the face of changing data distribution. The model should be able to maintain stable performance under different inputs.

[0183] Scalability: Test the performance of the model and system architecture when handling increasing workloads. Good scalability means that the system can maintain a stable service level as demand grows.

[0184] Specifically, the charging strategy generation model uses evaluation indicators to evaluate the performance of the classification model. The specific steps are as follows:

[0185] a) Performance monitoring and evaluation

[0186] After the model is deployed, its performance in the production environment is continuously monitored, including but not limited to key indicators such as accuracy, error rate, response time, etc. At the same time, A / B testing is carried out to compare the performance of new and old models or different model versions to ensure that the update brings a positive impact.

[0187] b) Data Feedback Loop

[0188] Collect new data: Continuously collect new user behavior data and billing results, and fine-tune the existing model based on the results to reflect the latest data characteristics.

[0189] c) Incremental Learning

[0190] Use new data or a small portion of historical data to update the model, reduce computing resource consumption, and speed up iteration.

[0191] In this embodiment, the charging policy generation device establishes a preset charging policy generation model and regularly trains the model so that the model can analyze user behavior data in real time, instantly capture changes in user needs, and dynamically adjust the charging policy.

[0192] It should be noted that the above examples are only used to understand the present application and do not constitute a limitation on the billing policy generation method of the present application. More simple transformations based on this technical concept are all within the scope of protection of the present application.

[0193] This application also provides a charging strategy generation device, please refer to Figure 6 , the charging strategy generating device comprises:

[0194] An acquisition module 10 is used to acquire the target user's call list data collected in real time and the user elements and billing dimension data stored in the cache database;

[0195] An extraction module 20, configured to input the target user call list data, the user elements and the billing dimension data into an updated preset billing strategy generation model for feature extraction to obtain a user data vector and a user demand vector;

[0196] The generation module 30 is used to determine the attention weight based on the user data vector and the preset billing strategy generation model, generate the billing strategy based on the attention weight, the user data vector and the user demand vector, and obtain the target billing strategy.

[0197] The charging policy generation device provided by the present application adopts the charging policy generation method in the above embodiment, which can solve the technical problem of charging policy generation. Compared with the prior art, the beneficial effects of the charging policy generation device provided by the present application are the same as the beneficial effects of the charging policy generation method provided by the above embodiment, and other technical features in the charging policy generation device are the same as the features disclosed in the above embodiment method, which will not be repeated here.

[0198] The present application provides a billing policy generation device, which includes: at least one processor; and a memory that is communicatively connected to the at least one processor; wherein the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the billing policy generation method in the above-mentioned embodiment 1.

[0199] Reference below Figure 7 , which shows a schematic diagram of the structure of a charging policy generation device suitable for implementing the embodiment of the present application. The charging policy generation device in the embodiment of the present application may include but is not limited to mobile terminals such as mobile phones, laptops, tablet computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PMPs (Portable Media Players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 7 The charging policy generation device shown is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0200] like Figure 7 As shown, the charging policy generating device may include a processing device 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM: Read Only Memory) 1002 or a program loaded from a storage device 1003 to a random access memory (RAM: Random Access Memory) 1004. In RAM1004, various programs and data required for the operation of the charging policy generating device are also stored. The processing device 1001, ROM1002, and RAM1004 are connected to each other through a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Generally, the following systems can be connected to the I / O interface 1006: an input device 1007 including, for example, a touch screen, a touch pad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; an output device 1008 including, for example, a liquid crystal display (LCD: Liquid Crystal Display), a speaker, a vibrator, etc.; a storage device 1003 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 1009. The communication device 1009 can allow the charging policy generation device to communicate with other devices wirelessly or wired to exchange data. Although the figure shows a charging policy generation device with various systems, it should be understood that it is not required to implement or have all the systems shown. More or fewer systems can be implemented or provided instead.

[0201] In particular, according to the embodiments disclosed in the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present application include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication device, or installed from a storage device 1003, or installed from a ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the method of the embodiment disclosed in the present application are executed.

[0202] The charging policy generation device provided by the present application adopts the charging policy generation method in the above embodiment, which can solve the technical problem of charging policy generation. Compared with the prior art, the beneficial effects of the charging policy generation device provided by the present application are the same as the beneficial effects of the charging policy generation method provided by the above embodiment, and other technical features in the charging policy generation device are the same as the features disclosed in the method of the previous embodiment, which will not be repeated here.

[0203] It should be understood that the various parts disclosed in this application can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0204] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0205] The present application provides a computer-readable storage medium having computer-readable program instructions (ie, computer programs) stored thereon, and the computer-readable program instructions are used to execute the charging policy generating method in the above-mentioned embodiment.

[0206] The computer-readable storage medium provided in the present application may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, systems or devices, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, 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 disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, system or device. The program code contained on the computer-readable storage medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination of the above.

[0207] The computer-readable storage medium may be included in the charging policy generating device; or may exist independently without being assembled into the charging policy generating device.

[0208] The above-mentioned computer-readable storage medium carries one or more programs. When the above-mentioned one or more programs are executed by the billing policy generation device, the billing policy generation device: obtains the target user call list data collected in real time and the user elements and billing dimension data stored in the cache database; inputs the target user call list data, the user elements and the billing dimension data into the updated preset billing policy generation model for feature extraction to obtain a user data vector and a user demand vector; determines the attention weight based on the user data vector and the preset billing policy generation model, generates a billing policy based on the attention weight, the user data vector and the user demand vector, and obtains a target billing policy.

[0209] Computer program code for performing the operations of the present application may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0210] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present application. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0211] The modules involved in the embodiments described in this application may be implemented by software or hardware, wherein the name of the module does not constitute a limitation on the unit itself in some cases.

[0212] The readable storage medium provided in the present application is a computer-readable storage medium, which stores computer-readable program instructions (i.e., computer programs) for executing the above-mentioned charging policy generation method, and can solve the technical problem of charging policy generation. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in the present application are the same as the beneficial effects of the charging policy generation method provided in the above-mentioned embodiment, and will not be repeated here.

[0213] The present application also provides a computer program product, including a computer program, which implements the steps of the above-mentioned charging policy generation method when executed by a processor.

[0214] The computer program product provided by this application can solve the technical problem of generating charging policies. Compared with the prior art, the beneficial effects of the computer program product provided by this application are the same as the beneficial effects of the charging policy generating method provided by the above embodiment, which will not be described in detail here.

[0215] The above descriptions are only some embodiments of the present application, and are not intended to limit the patent scope of the present application. All equivalent structural changes made using the contents of the present application specification and drawings under the technical concept of the present application, or direct / indirect applications in other related technical fields are included in the patent protection scope of the present application.

Claims

1. A method for generating a charging strategy, characterized in that: The charging strategy generation method comprises: Obtain the target user's call list data collected in real time and the user elements and billing dimension data stored in the cache database; Inputting the target user call list data, the user elements and the billing dimension data into the updated preset billing strategy generation model for feature extraction to obtain a user data vector and a user demand vector; Based on the user data vector and the preset billing strategy generation model, an attention weight is determined, and a billing strategy is generated based on the attention weight, the user data vector and the user demand vector to obtain a target billing strategy.

2. The method for generating a charging policy according to claim 1, wherein: The step of determining the attention weight based on the user data vector and the preset charging strategy generation model, generating the charging strategy based on the attention weight, the user data vector and the user demand vector, and obtaining the target charging strategy comprises: Based on the preset parameters of the attention mechanism layer of the preset billing strategy generation model and the user data vector, generate an attention score for each vector in the user data vector, wherein the preset parameters are obtained based on the initial parameter optimization; Normalizing the attention score to obtain an attention weight; Combining the attention weight, the user data vector, and the user demand feature vector through addition pooling to obtain a pooled first data vector; Performing user consumption behavior learning on the first data vector to obtain a second data vector including an updated user consumption pattern; Interacting all features in the second data vector to obtain an interacted third data vector; A target strategy corresponding to the third data vector is generated through a preset mapping relationship between the preset strategy and the data vector.

3. The method for generating a charging policy according to claim 1, wherein: The charging dimension data includes the first personalized dimension or the second personalized dimension, and the step of obtaining the target user call list data collected in real time and the user elements and charging dimension data stored in the cache database includes: If the user terminal does not initiate a user demand, obtain the user usage record within a preset time, extract information through the historical user call list data in the user usage record, obtain the user factor, and determine the first personalized dimension obtained based on the analysis of the user usage record; If the user terminal initiates the user demand, a second personalized dimension is determined based on the user demand, and the user usage record within a preset time is obtained, and information is extracted through historical user call list data in the user usage record to obtain user factors; The user elements and the charging dimension data are stored in a cache database.

4. The method for generating a charging policy according to claim 3, wherein: After the step of storing the user element and the billing dimension data in a cache database, the following steps are included: Calculate the mean and standard deviation of the numerical data in the historical user call list data, the user elements and the billing dimension data; Based on the mean and the standard deviation, the numerical data is scaled to a preset range to obtain the scaled historical user call list data, the user elements and the billing dimension data; Obtaining a charging history strategy corresponding to the historical user call list data, and using the historical user call list data, the user element and the charging dimension data as sample data, wherein the historical strategy is a sample label corresponding to the sample data; Obtaining an initial billing policy generation model; The sample data and the sample labels are input into the initial charging policy generation model for iterative training to obtain a preset charging policy generation model that meets the accuracy requirements.

5. The method for generating a charging policy according to claim 4, wherein: The step of inputting the sample data and the sample label into the initial charging policy generation model for iterative training to obtain a preset charging policy generation model that meets the accuracy requirement includes: Based on the input layer of the initial charging policy generation model, converting the sample data into multi-dimensional user-related data, and outputting the user-related data to the embedding layer of the initial charging policy generation model; Based on the embedding layer, the user data and user needs in the user-related data are parsed, the parsed user-related data is converted into a preset base vector, and the preset base vector corresponding to the user data and the preset base vector corresponding to the user need are spliced ​​respectively, and the spliced ​​user data vector and user need vector are input into the attention mechanism layer of the initial billing strategy generation model; Based on the attention mechanism layer and the initial parameters of the attention mechanism layer, the user data vector and the user demand feature vector are combined by addition pooling, and the weighted first sample vector is input into the memory network layer of the initial charging strategy generation model; Based on the memory network layer and the initial weight matrix and initial bias vector of the memory network layer, the first sample vector is subjected to user consumption behavior learning, and the learned second sample vector including the updated user consumption pattern is input into the interaction layer of the initial charging strategy generation model; Based on the interaction layer, all features in the second sample vector are interacted, and the third sample vector after interaction is input into the decision output layer of the initial charging strategy generation model; Based on the decision output layer, performing strategy mapping on the third sample vector to obtain an initial charging strategy; Based on the sample labels and the initial charging strategy, the initial parameters, initial weight matrix and initial bias vector are optimized to obtain a preset charging strategy generation model that meets the accuracy requirements after optimization.

6. The method for generating a charging policy according to claim 4, wherein: The step of inputting the sample data and the sample label into the initial charging policy generation model for iterative training to obtain a preset charging policy generation model that meets the accuracy requirement includes: Storing the preset charging policy generation model in a shared memory library; Based on a preset timing time, the preset charging policy generation model stored in the shared memory library is updated to obtain an updated preset charging policy generation model.

7. A charging strategy generating device, characterized in that: The device comprises: An acquisition module is used to acquire the target user's call list data collected in real time and the user elements and billing dimension data stored in the cache database; An extraction module, used for inputting the target user call list data, the user elements and the billing dimension data into the updated preset billing strategy generation model for feature extraction to obtain a user data vector and a user demand vector; A generation module is used to determine the attention weight based on the user data vector and the preset billing strategy generation model, and generate a billing strategy based on the attention weight, the user data vector and the user demand vector to obtain a target billing strategy.

8. A charging policy generating device, characterized in that: The device comprises: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the charging policy generating method according to any one of claims 1 to 6.

9. A storage medium, characterized in that: The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, the steps of the charging policy generation method according to any one of claims 1 to 6 are implemented.

10. A computer program product, characterized in that The computer program product comprises a computer program, and when the computer program is executed by a processor, the steps of the charging policy generating method according to any one of claims 1 to 6 are implemented.

Images