Business execution method and device

Abstract

The embodiments of this specification provide a service execution method and a device. The service execution method includes: first, obtaining an execution request for a target service, where the execution request carries pre-investment data and service target data for executing the target service; then, constructing a target service execution model with the pre-investment data and service target data as constraint conditions; and then, using the target service execution model to determine an execution strategy for executing the target service. By establishing a target service execution model with the pre-investment data and service target data carried in the execution request of the target service as constraint conditions, the service is unifiedly planned, and the efficiency of determining the execution strategy is improved to a certain extent, enabling a unified and optimal execution strategy during service execution.

Description

Technical Field

[0001] The embodiments of the present specification relate to the field of computer technology, and particularly to a service execution method. Background Art

[0002] With the rapid development of computer technology, various advertising display methods emerge in an endless stream. In the field of advertising display, online display advertising is very popular. In the scenario of online display advertising, advertisers usually obtain exposure opportunities through real-time bidding. The most common requirement of advertisers on most advertising platforms is to maximize the value of the won traffic under the constraints of budget and certain key performance indicators (KPIs), such as maximizing the number of clicks under the constraints of budget and click cost. The placement requirements of each advertiser vary greatly in three dimensions: marketing objectives (such as clicks, exposures), types of KPI constraints (such as upper bound of click cost, lower bound of click-through rate), and the number of KPI constraints.

[0003] In the prior art, linear interpolation is used to adjust parameters. On the premise that the parameter competition environment throughout the day is relatively stable, the parameters are adjusted in combination with the placement results of the previous adjustment period and the remaining adjustment time. However, the above method is limited to the placement requirements with a relatively stable parameter competition environment, lacks generality, and it is difficult to achieve the optimal placement result. Therefore, an effective solution is urgently needed to solve the above problems. Summary of the Invention

[0004] In view of this, the embodiments of the present specification provide a service execution method. One or more embodiments of the present specification also relate to a service execution device, a computing device, a computer-readable storage medium, and a computer program to solve the technical defects existing in the prior art.

[0005] According to the first aspect of the embodiments of the present specification, a service execution method is provided, including:

[0006] Obtain an execution request for a target service, where the execution request carries pre-investment data and service target data for executing the target service;

[0007] Construct a target service execution model with the pre-investment data and service target data as constraint conditions;

[0008] Use the target service execution model to determine an execution strategy for executing the target service.

[0009] Optionally, constructing a target service execution model with the pre-investment data and service target data as constraint conditions includes:

[0010] Construct a linear programming model with the pre-invested execution data and business objective data as constraints, where the linear programming model represents the constraint relationship between the business objective data and the pre-invested data;

[0011] Use the linear programming model to construct an initial business execution model;

[0012] According to the preset adjustment rules, adjust at least one business execution parameter of the initial business execution model to obtain the target business execution model.

[0013] Optionally, using the linear programming model to construct an initial business execution model includes:

[0014] Identify the data type of the business objective data;

[0015] According to the data type of the business objective data, divide the business objective data into first-class data and second-class data. The first-class data is related to the pre-invested data, and the second-class data is not related to the pre-invested data;

[0016] According to the first-class data and the second-class data, use the linear programming model to construct an initial business execution model.

[0017] Optionally, according to the preset adjustment rules, adjusting at least one business execution parameter of the initial business execution model to obtain the target business execution model includes:

[0018] Obtain the execution status of historical business;

[0019] According to the execution status of historical business, adjust the initial values of at least one business execution parameter to obtain the target values of at least one business execution parameter;

[0020] According to the target values of at least one business execution parameter, update the initial business execution model to obtain the target business execution model.

[0021] Optionally, according to the execution status of historical business, adjusting the initial values of at least one business execution parameter to obtain the target values of at least one business execution parameter includes:

[0022] According to the execution status of historical business and the initial values of at least one business execution parameter, use an evaluation function to determine the evaluation result corresponding to the initial values of at least one business execution parameter;

[0023] According to the evaluation result, use a policy function to determine the adjustment policy for the initial values of at least one business execution parameter;

[0024] According to the adjustment policy, adjust the initial values of at least one business execution parameter to obtain the target values of at least one business execution parameter.

[0025] Optionally, according to the execution status of historical services, adjust the initial values of at least one service execution parameter to obtain the target values of at least one service execution parameter, including:

[0026] Calculate the difference between the execution status of the historical service and the target execution status corresponding to the preset historical service;

[0027] Based on the difference, adjust the initial values of at least one service execution parameter to obtain the target values of at least one service execution parameter.

[0028] Optionally, according to the execution status of historical services, adjust the initial values of at least one service execution parameter to obtain the target values of at least one service execution parameter, including:

[0029] Measure the initial values of at least one service execution parameter according to the execution status of the historical service and a preset measurement rule to obtain a measurement value;

[0030] In the case where the measurement value is less than the measurement threshold, adjust the initial values of at least one service execution parameter, and perform the step of measuring the initial values of at least one service execution parameter according to the execution status of the historical service and a preset measurement rule;

[0031] In the case where the measurement value is greater than or equal to the measurement threshold, determine the initial values of at least one service execution parameter as the target values of at least one service execution parameter.

[0032] Optionally, according to the execution status of historical services, adjust the initial values of at least one service execution parameter to obtain the target values of at least one service execution parameter, including:

[0033] Obtain a preset time value;

[0034] At each interval of the time value, adjust the initial values of at least one service execution parameter according to the execution status of the historical service to obtain the target values of at least one service execution parameter.

[0035] Optionally, with the pre-investment data and business target data as constraint conditions, construct a linear programming model, including:

[0036] With the pre-investment data as the first constraint condition, construct a pre-investment execution algorithm that conforms to linear programming;

[0037] With the business target data as the second constraint condition, construct a business target execution algorithm that conforms to linear programming;

[0038] According to the pre-investment execution algorithm and the business target execution algorithm, construct a linear programming model.

[0039] According to the second aspect of the embodiments of the present specification, a service execution device is provided, including:

[0040] An acquisition module, configured to acquire an execution request for a target service, where the execution request carries pre-investment data and service target data for executing the target service;

[0041] A construction module, configured to construct a target service execution model with the pre-investment data and the service target data as constraint conditions;

[0042] A determination module, configured to determine an execution strategy for executing the target service by using the target service execution model.

[0043] Optionally, the construction module is further configured to:

[0044] Construct a linear programming model with the pre-investment execution data and the service target data as constraint conditions, where the linear programming model represents the constraint relationship between the service target data and the pre-investment data;

[0045] Construct an initial service execution model by using the linear programming model;

[0046] Adjust at least one service execution parameter of the initial service execution model according to a preset adjustment rule to obtain the target service execution model.

[0047] Optionally, the construction module is further configured to:

[0048] Identify the data type of the service target data;

[0049] Divide the service target data into first-class data and second-class data according to the data type of the service target data, where the first-class data is related to the pre-investment data and the second-class data is not related to the pre-investment data;

[0050] Construct an initial service execution model by using the linear programming model according to the first-class data and the second-class data.

[0051] Optionally, the construction module is further configured to:

[0052] Obtain the execution status of historical services;

[0053] Adjust the initial values of at least one service execution parameter according to the execution status of historical services to obtain the target values of at least one service execution parameter;

[0054] Update the initial service execution model according to the target values of at least one service execution parameter to obtain the target service execution model.

[0055] Optionally, the construction module is further configured to:

[0056] Determine an evaluation result corresponding to an initial value of at least one service execution parameter by using an evaluation function according to an execution state of a historical service and an initial value of at least one service execution parameter;

[0057] Determine an adjustment strategy for an initial value of at least one service execution parameter by using a policy function according to the evaluation result;

[0058] Adjust the initial value of at least one service execution parameter according to the adjustment strategy to obtain a target value of at least one service execution parameter.

[0059] Optionally, the construction module is further configured to:

[0060] Calculate a difference between an execution state of a historical service and a target execution state corresponding to a preset historical service;

[0061] Based on the difference, adjust the initial value of at least one service execution parameter to obtain a target value of at least one service execution parameter.

[0062] Optionally, the construction module is further configured to:

[0063] Measure an initial value of at least one service execution parameter according to a preset measurement rule according to an execution state of a historical service to obtain a measurement value;

[0064] In a case where the measurement value is less than a measurement threshold, adjust the initial value of at least one service execution parameter, and perform a step of measuring the initial value of at least one service execution parameter according to a preset measurement rule according to an execution state of a historical service;

[0065] In a case where the measurement value is greater than or equal to the measurement threshold, determine the initial value of at least one service execution parameter as the target value of at least one service execution parameter.

[0066] Optionally, the construction module is further configured to:

[0067] Obtain a preset time value;

[0068] At each interval of the time value, adjust an initial value of at least one service execution parameter according to an execution state of a historical service to obtain a target value of at least one service execution parameter.

[0069] Optionally, the construction module is further configured to:

[0070] Construct a pre-investment execution algorithm that conforms to linear programming with pre-investment data as a first constraint condition;

[0071] Construct a business objective execution algorithm that conforms to linear programming with business objective data as a second constraint condition;

[0072] Construct a linear programming model according to the pre-investment execution algorithm and the business objective execution algorithm.

[0073] According to the third aspect of the embodiments of the present specification, a computing device is provided, including:

[0074] A memory and a processor;

[0075] The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions:

[0076] Obtain an execution request for a target business, where the execution request carries pre-investment data and business objective data for executing the target business;

[0077] Construct a target business execution model with the pre-investment data and the business objective data as constraint conditions;

[0078] Use the target business execution model to determine an execution strategy for executing the target business.

[0079] According to the fourth aspect of the embodiments of the present specification, a computer-readable storage medium is provided, which stores computer-executable instructions, and when the instructions are executed by a processor, the steps of any one of the business execution methods are implemented.

[0080] According to the fifth aspect of the embodiments of the present specification, a computer program is provided, where when the computer program is executed on a computer, the computer is made to execute the steps of the above-mentioned business execution method.

[0081] A business execution method provided in the present specification: Obtain an execution request for a target business, where the execution request carries pre-investment data and business objective data for executing the target business; Construct a target business execution model with the pre-investment data and the business objective data as constraint conditions; Use the target business execution model to determine an execution strategy for executing the target business. By establishing a target business execution model with the pre-investment data and the business objective data carried in the execution request of the target business as constraint conditions, the business is unifiedly planned, and the efficiency of determining the execution strategy is improved to a certain extent, so that there can be a unified and optimal execution strategy during business execution. Description of the Drawings

[0082] Figure 1 is a flowchart of a business execution method provided by an embodiment of the present specification;

[0083] Figure 2 is a process flowchart of a business execution method provided by an embodiment of the present specification;

[0084] Figure 3 is a schematic structural diagram of a business execution device provided by an embodiment of the present specification;

[0085] Figure 4 It is a structural block diagram of a computing device provided by an embodiment of this specification. Detailed implementation manners

[0086] In the following description, many specific details are set forth in order to provide a thorough understanding of this specification. However, this specification can be implemented in many other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the connotation of this specification. Therefore, this specification is not limited by the specific implementations disclosed below.

[0087] The terms used in one or more embodiments of this specification are only for the purpose of describing specific embodiments, and are not intended to limit one or more embodiments of this specification. The singular forms of "a", "the", and "said" used in one or more embodiments of this specification and the appended claims are also intended to include the plural forms, unless the context clearly indicates otherwise. It should also be understood that the term "and / or" used in one or more embodiments of this specification refers to and includes any or all possible combinations of one or more of the associated listed items.

[0088] It should be understood that although the terms first, second, etc. may be used in one or more embodiments of this specification to describe various information, such information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of one or more embodiments of this specification, the first may also be referred to as the second, and similarly, the second may also be referred to as the first. Depending on the context, the word "if" as used herein may be interpreted as "when" or "while" or "in response to determining".

[0089] First, the noun terms related to one or more embodiments of this specification are explained.

[0090] Linear programming (LP), is an important branch in operations research that has been studied early, developed rapidly, widely applied, and has relatively mature methods. It is a mathematical method to assist people in scientific management. It studies the mathematical theory and methods for the extreme value problems of linear objective functions under linear constraint conditions.

[0091] Real Time Bidding (RTB) is a bidding technology that uses third-party technologies to evaluate and bid on the display behaviors of each user on millions of websites.

[0092] Key Performance Indicator (KPI): It refers to an objective quantitative management indicator that measures process performance by setting, sampling, calculating, and analyzing key parameters, such as display, click-through rate, cost per add to cart, minimum click-through rate, etc.

[0093] Constrained bidding problem: In the case of budget and multiple KPI constraints, it helps advertisers optimize their advertising placement effects during the RTB process.

[0094] Proportion Integral Differential (PID) algorithm: The control algorithm combines three links of proportion, integral, and differential.

[0095] Parameter regulation service: The bid for each page view (PV) in the automatic bidding plan depends on the bid parameters.

[0096] Next, a brief introduction to the basic concept of the business execution method provided in this specification is given.

[0097] In the online display advertising scenario, advertisers usually obtain exposure opportunities through real-time bidding. On most advertising platforms, the most common need of advertisers is to maximize the value of the acquired traffic under budget and certain KPI constraints (such as maximizing the click-through rate under budget and click cost constraints). The advertising placement requirements of each advertiser vary greatly in three dimensions: marketing objectives (such as clicks, exposures), KPI constraint types (such as upper bound of click cost, lower bound of click-through rate), and the number of KPI constraints.

[0098] In the prior art, most methods use linear interpolation to regulate parameters. Based on the premise that the parameter competition environment throughout the day is relatively stable, parameters are regulated by combining the placement results of the previous regulation period and the remaining regulation time; or use the PID algorithm to regulate parameters, that is, manually set the consumption target and various index target values for each time period before the planned placement, and then use the PID algorithm to regulate parameters.

[0099] However, the above methods are limited to a specific advertising placement requirement, lack generality, or it is difficult to achieve the optimal placement result. Therefore, a business execution method that can solve the above problems is proposed in this specification.

[0100] In this specification, a business execution method is provided. This specification also relates to a business execution device, a computing device, and a computer-readable storage medium, which will be described in detail one by one in the following embodiments.

[0101] Figure 1The flowchart of a service execution method provided according to an embodiment of this specification is shown, which specifically includes the following steps.

[0102] Step 102: Obtain an execution request for a target service, where the execution request carries pre-investment data and service target data for executing the target service.

[0103] Specifically, the service can be an advertising service, a procurement service, a handling service, etc., and this application does not limit it; the target service refers to the service for which an execution strategy needs to be determined; the pre-investment data refers to the data that the user or customer expects to invest when executing the target service, such as the expected resource value to be invested, the expected human and material resources, etc.; the service target data refers to the index data that is expected to be achieved when the service is executed, and the service target data can include at least one of cost per mille (CPM), cost per click (CPC), cost per action (CPA), return on investment (ROI), click-through rate (CTR), and conversion per impression (CPI).

[0104] In practical applications, the user fills in pre-investment data, service target data, etc. for the service that the user needs to execute through the client or platform, and after clicking the confirmation button, the local area will receive the execution request for the service that the user needs to execute, that is, the execution request for the target service. Among them, the execution request also carries data such as the pre-investment data and service target data filled in by the user.

[0105] For example, the local area receives an execution request from the user for the target procurement service, where the execution request carries the pre-investment data "100 resource values" and the service target data "cost per action 1 resource value" for the target procurement service.

[0106] In this embodiment, by obtaining the execution request carrying the pre-investment data and service target data for executing the target service, the pre-investment data and service target data can be quickly determined, laying a foundation for subsequently determining the execution strategy of the target service according to the pre-investment data and service target data, which is beneficial to improving the efficiency of determining the execution strategy and further improving the efficiency of service execution.

[0107] Step 104: Construct a target service execution model with the pre-investment data and the service target data as constraint conditions.

[0108] On the basis of obtaining an execution request carrying pre-investment data and business target data for executing a target business, further, perform linear programming according to the pre-investment data and the business target data to construct a target business execution model.

[0109] Specifically, a constraint condition refers to various restrictive conditions on the value range of design variables in an optimization design. The pre-investment data and each business target data can be used as constraint conditions. Linear programming is a method for scientific management in the process of business execution. A business execution model refers to a model that takes business target data and pre-investment data as inputs and an execution strategy as an output. That is, the business execution model represents the relationship between business target data, pre-investment data, and the execution strategy. A target business execution model refers to a business execution model with parameters being specified values.

[0110] In practical applications, after obtaining the pre-investment data and business target data for executing the target business, it is necessary to use the pre-investment data and business target data as constraint conditions to further construct a target business execution model: The pre-investment data and business target data can be used as one constraint condition to construct a target business execution model; the pre-investment data can be used as one constraint condition and the business target data can be used as another constraint condition to construct a target business execution model; the pre-investment data can be used as one constraint condition and the pre-investment data and business target data can be used as another constraint condition to construct a target business execution model; the business target data can be used as one constraint condition and the pre-investment data and business target data can be used as another constraint condition to construct a target business execution model. This application does not make any limitations in this regard.

[0111] When constructing the target business execution model, for the effectiveness and accuracy of the target business execution model, and to further improve the accuracy and effectiveness of the finally determined execution strategy, it is necessary to determine and establish a linear programming model, construct an initial business execution model according to the linear programming model, and then adjust the business execution parameters in the initial business execution model according to a preset adjustment rule to obtain the target business execution model. That is, the specific implementation process of constructing the target business execution model with the pre-investment data and the business target data as constraint conditions can be as follows:

[0112] Construct a linear programming model with the pre-investment execution data and the business target data as constraint conditions, and the linear programming model represents the constraint relationship between the business target data and the pre-investment data;

[0113] Use the linear programming model to construct an initial business execution model;

[0114] According to a preset adjustment rule, adjust at least one business execution parameter of the initial business execution model to obtain the target business execution model.

[0115] Specifically, a linear programming model refers to a model that represents the constraint relationship between business objective data and pre-investment data; business execution parameters refer to the parameters in the initial business execution model or the target business execution model.

[0116] In practical applications, it is necessary to first construct a linear programming model that represents the constraint relationship between business objective data and pre-investment data with the pre-investment execution data and business objective data as constraint conditions; then construct an initial business execution model based on the linear programming model; then determine at least one business execution parameter of the initial business execution model according to the preset adjustment rules, that is, adjust at least one business execution parameter of the initial business execution model. The initial business execution model after adjusting at least one business execution parameter is the target business execution model.

[0117] For example, the pre-investment execution data is a, and the business objective data are b1 and b2. A linear programming model is constructed based on the pre-investment execution data a and the business objective data b: max z = b1 + b2 and hb1 + gb2 ≤ a, where z represents the value when the business objective data are b1 and b2, and h and g are two linear programming parameters; then an initial business execution model is constructed according to the linear programming model: s = e1a + e2h + e2g, where e1 and e2 are business execution parameters and s is the optimal output value of the business; adjust e1 and e2 according to the preset adjustment rules, and after inputting the adjusted e1 and e2 into the initial business execution model, the target business execution model is obtained.

[0118] In order to make full use of the pre-investment data and business objective data, improve the execution strategy output by the target business execution model and the pre-investment data and business objective data, and improve the superiority of the execution strategy, while reducing redundant data processing, the pre-investment data and business objective data can be used as constraint conditions respectively, and then a linear programming model can be constructed. That is, taking the pre-investment data and the business objective data as constraint conditions, the specific implementation process of constructing a linear programming model can be as follows:

[0119] Taking the pre-investment data as the first constraint condition, construct a pre-investment execution algorithm that conforms to linear programming;

[0120] Taking the business objective data as the second constraint condition, construct a business objective execution algorithm that conforms to linear programming;

[0121] Construct the linear programming model according to the pre-investment execution algorithm and the business objective execution algorithm.

[0122] Specifically, the first constraint condition refers to the constraint condition with pre-investment data as the constraint; the pre-investment execution algorithm refers to the algorithm with pre-investment data as the input and output; the second constraint condition refers to the constraint condition with business target data as the constraint; the business target execution algorithm refers to the algorithm with business target data as the input.

[0123] In practical applications, in order to avoid exceeding the pre-investment data of the user or the target business, the pre-investment data can be used as the first constraint condition, that is, the pre-investment data is used as the upper bound of the input data, and a pre-investment execution algorithm that conforms to linear programming is constructed, as shown in Equation 1.

[0124] Σ i c i x i ≤B (Equation 1)

[0125] In Equation 1, i represents the effect of the target business; c i is the input data required to achieve the effect i; x i is a binary index indicating whether the effect i is achieved; B is the pre-investment data.

[0126] Furthermore, in order to avoid exceeding the business target data of the user or the target business, the business target data can be used as the second constraint condition, that is, the business target data is used as the upper bound, and a business target execution algorithm that conforms to linear programming is constructed, as shown in Equation 2.

[0127] (Σ i c i I CRj +q ij (1-I CRj )x i ) / (Σ i P ij x i )≤K j (Equation 2)

[0128] In Equation 2, i represents the effect of the target business; c i is the input data required to achieve the effect i, q ij and P ij can be any performance indicator or constant, that is, a parameter; I CRj is an index function indicating whether the jth business target data is related to the pre-investment data; x i is a binary index indicating whether the effect i is achieved; K j represents the upper bound of the jth business target data.

[0129] After obtaining the pre-investment execution algorithm and the business objective execution algorithm, further, based on the common objective of the target business - maximizing the total value obtained, a linear programming model is constructed according to the pre-investment execution algorithm and the business objective execution algorithm, as shown in Equation 3.

[0130] maxΣ i v i x i

[0131] s.tΣ i c i x i ≤B

[0132]

[0133]

[0134]

[0135] In Equation 3, i represents the effect of the target business; v i represents the value of effect i; c i is the data required to achieve effect i; x i is a binary indicator representing whether effect i is achieved; q ij and P ij can be any performance metric or constant, that is, a parameter; I CRj is an index function indicating whether the j-th business objective data is related to the pre-investment data; K j represents the upper bound of the j-th business objective data; B is the pre-investment data.

[0136] For example, the target business is an advertising business, the pre-investment data for this advertising business is 500, the business objective data includes a cost per click of 20 and a click-through rate of 80, the effect of this advertising business is the number of exposures, and an investment execution algorithm is established with 500 as the first constraint condition; with the 20 and 80 as the second constraint conditions, a business objective execution algorithm that conforms to linear programming is constructed, and the linear programming model constructed according to the pre-investment execution algorithm and the business objective execution algorithm is as shown in Equation 4.

[0137] maxΣ i v i x i

[0138] s.tΣ i c i x i ≤500

[0139] (Σ i c i I CR1 +qi1 (1 - I CR1 )x i ) / (Σ i P i1 x i ) ≤ 20

[0140] (Σ i c i I CR2 + q i2 (1 - I CR2 )x i ) / (Σ i P i2 x i ) ≤ 80

[0141]

[0142]

[0143] In Equation 4, i represents the number of exposures of the advertising service; v i represents the value of the i-th exposure; c i is the input data required to achieve the i-th effect; x i is a binary indicator representing whether the i-th exposure can be achieved; q i1 and P i1 are parameters of the cost per click; q i2 and P i2 are constants of the click-through rate; I CR1 is an indicator function representing whether the first business target data is related to the pre-input data; I CR2 is an indicator function representing whether the second business target data is related to the pre-input data.

[0144] In one or more embodiments of the present embodiment, in order to improve the accuracy and reliability of the target business execution model, first, the reliability of the initial business execution model needs to be improved. The business target data can be divided according to the type of the business target data, and then the initial business execution model can be constructed by using the linear programming model. That is, the specific implementation process of constructing the initial business execution model by using the linear programming model can be as follows:

[0145] Identify the data type of the business target data;

[0146] According to the data type of the business target data, divide the business target data into the first type of data and the second type of data. The first type of data is related to the pre-input data, and the second type of data is not related to the pre-input data;

[0147] Construct the initial service execution model according to the first type of data and the second type of data by using the linear programming model.

[0148] Specifically, the data types include data types related to pre-investment data and data types unrelated to pre-investment data; the first type of data includes service target data related to pre-investment data; the second type of data includes service target data unrelated to pre-investment data.

[0149] In practical applications, there may be multiple service target data. Classify them according to the data types of each service target data, that is, classify them according to whether the service target data is related to pre-investment data. Classify the service target data related to pre-investment data as the first type of data, and classify the service target data unrelated to pre-investment data as the second type of data; further construct the initial service execution model according to the first type of data and the second type of data in combination with the linear programming model.

[0150] For example, there are three service target data D1, D2, and D3. Among them, D1 and D3 are both related to pre-investment data, and D2 is unrelated to pre-investment data. Then classify D1 and D3 as the first type of data, classify D2 as the second type of data, and then construct the initial service execution model according to the first type of data and the second type of data.

[0151] Preferably, constructing the initial service execution model according to the first type of data and the second type of data may be: determine the coefficient of the first type of data according to the first type of data, determine the coefficient of the second type of data according to the second type of data, divide the coefficient of the second type of data by the coefficient of the first type of data, and then determine the quotient of the coefficient of the second type of data and the coefficient of the first type of data in combination with the linear programming model, and construct the initial service execution model, as shown in Equation 5.

[0152] b * i =P * NCR / P * CR =w * 0v i -Σ j w * j (q ij (1-I CRj )-K j P ij ) (Equation 5)

[0153] In Equation 5, b * i is the optimal input data; P * NCR is the coefficient of the second type of data; P *CR is the coefficient of the first type of data; w * 0 and w * j are business execution parameters; v i represents the value of effect i; q ij and P ij can be any performance metric or constant, that is, a parameter; I CRj is an index function indicating whether the j-th business target data is related to the pre-investment data; K j represents the upper bound of the j-th business target data.

[0154] Thus, according to Equation 3, two constants q ij and P ij can be determined. Then, according to q ij and P ij the quotient of the coefficient of the second type of data and the coefficient of the first type of data is determined, thereby obtaining the initial business execution model. That is, substituting q ij and P ij into Equation 5, the initial business execution model is obtained.

[0155] Since the execution environment of the business changes over time, that is, the business execution parameters also change accordingly. Before determining the execution strategy using the target business execution model, if the initial business execution model is used, that is, the business execution parameters remain unchanged, it will lead to a significant reduction in the accuracy and effectiveness of the execution strategy. Therefore, in order to improve the accuracy and effectiveness of the finally determined execution strategy, it is necessary to determine the business execution parameters in the current environment. That is, at least one business execution parameter in the initial business execution model needs to be adjusted according to the preset adjustment rules, and then the target business execution model is determined based on the at least one adjusted business execution parameter. That is, according to the preset adjustment rules, at least one business execution parameter of the initial business execution model is adjusted to obtain the target business execution model. The specific implementation process is as follows:

[0156] Obtain the execution status of historical business;

[0157] According to the execution status of the historical business, adjust the initial value of the at least one business execution parameter to obtain the target value of the at least one business execution parameter;

[0158] According to the target value of the at least one business execution parameter, update the initial business execution model to obtain the target business execution model.

[0159] Specifically, a historical service refers to a service that has been executed; the execution status includes the execution effect of the historical service; a service execution parameter refers to a parameter in an initial service execution model or a target service execution model; the initial value of a service execution parameter refers to the value corresponding to the service execution parameter before adjustment.

[0160] In practical applications, in order to better adjust service execution parameters, it is necessary to obtain the execution status of historical services, and based on the execution status of historical services, such as execution effects, etc., adjust the initial values of at least one service execution parameter, and then obtain the target values of at least one service execution parameter. Adjust the service execution parameters in the initial service execution model to the target values, that is, obtain the updated initial service execution model, which is also the target service execution model.

[0161] For example, the initial service execution model has two service execution parameters: PA1 and PA2, where the initial value of PA1 is pa11 and the initial value of PA2 is pa21. Obtain the execution status of the historical service, and then based on the execution status, adjust the initial value pa11 of PA1 to the target value pa12, and adjust the initial value pa21 of PA2 to the target value pa22; then use the target value pa12 to replace the initial value pa11 in the initial service execution model, and use the target value pa22 to replace the initial value pa21 in the initial service execution model, to obtain the updated initial service execution model, and the updated initial service execution model is the target service execution model.

[0162] In one or more embodiments of this embodiment, according to the execution status of historical service C, based on a reinforcement learning model, such as the Actor-Critic (AC) algorithm, the initial values of at least one service execution parameter can be adjusted. In this way, the values of service execution parameters can converge quickly, so as to obtain the target values of service execution parameters, and further make the execution strategy better. That is, the specific implementation process of adjusting the initial values of the at least one service execution parameter according to the execution status of the historical service to obtain the target values of the at least one service execution parameter can be:

[0163] According to the execution status of the historical service and the initial values of the at least one service execution parameter, use an evaluation function to determine the evaluation result corresponding to the initial values of the at least one service execution parameter;

[0164] According to the evaluation result, use a policy function to determine the adjustment strategy for the initial values of the at least one service execution parameter;

[0165] According to the adjustment strategy, adjust the initial values of the at least one service execution parameter to obtain the target values of the at least one service execution parameter.

[0166] Specifically, the policy function can be an actor function, i.e., an Actor function, which is responsible for interacting with the execution state of historical services and generating an action, i.e., a policy result; the evaluation function can be a critic function, i.e., a Critic function, which is a value function and is responsible for evaluating the performance of the Actor, i.e., the performance of the initial values of at least one service execution parameter.

[0167] In practical applications, after obtaining the execution state of historical services, the initial values of at least one service execution parameter can be evaluated according to the Critic function to obtain the evaluation results of the initial values of each service execution parameter. Further, the evaluation results of the initial values of each service execution parameter are respectively input into the Actor function to obtain the adjustment strategy for the initial values of at least one service execution parameter. Further, the initial values of each service execution parameter are adjusted according to the adjustment strategy. Then, it is necessary to execute the step of evaluating the adjusted initial values of at least one service execution parameter according to the Critic function again until the initial values are adjusted according to the adjustment strategy and the adjusted initial values no longer change or converge, and the initial values at this time are determined as the target values, i.e., the target values of the at least one service execution parameter are obtained.

[0168] For example, according to the execution state of historical services, the adjustment of the initial values of at least one service execution parameter based on the Actor-Critic (AC) algorithm can be as follows:

[0169] Initialize a random process ε of dimension M + 1;

[0170] Initialize a replay memory M with a capacity of N;

[0171] Initialize an actor π with weights θ θ ;

[0172] Initialize a critic Q with weights θ η ;

[0173] Set the batch size to BS;

[0174] Let be the optimal service execution parameter vector (w * 0,..., w * M );

[0175] Let R * be the theoretical optimal result;

[0176] When not converging, execute:

[0177] Randomly select an advertising campaign and simulate SPA;

[0178] Let

[0179] At time step 1, use to place a bid and obtain a reward r1;

[0180] Let R = r1;

[0181] When t = 2 to T, execute:

[0182] Observe the state s t ;

[0183] Obtain the action vector

[0184] Let

[0185] At time step t, use to place a bid and obtain a reward r t ;

[0186] Let R = R + r t and V = 0;

[0187] When τ = 2 to T, execute:

[0188] At time step τ, use to place a bid and obtain a reward r τ ;

[0189] Let V = V + r τ

[0190] End;

[0191] Calculate the penalty p for the KPI constraint j j ;

[0192] Let

[0193] Store in the memory M;

[0194] Extract the tuple BS from M

[0195] Update the critic Q by reducing the loss η :

[0196]

[0197] Update the actor π by policy gradient θ :

[0198]

[0199] End;

[0200] End.

[0201] where, V is the sum of rewards r τ and also the sum of rewards within the time period 2 - T; is a guiding signal used to distinguish which adjustment will make the service execution parameter better, and R is the current result corresponding to R * ; is the update function of the critic Q η ; is the update function of the actor π θ , that is, the policy gradient function.

[0202] It can be seen that according to the above method, the service execution parameters w * 0 and w * j can be determined, and the target values of w * 0 and w * j are input into the initial service execution model, then the target service execution model is obtained.

[0203] In one or more embodiments of this embodiment, according to the execution state of the historical service, the specific implementation process of adjusting the initial value of the at least one service execution parameter to obtain the target value of the at least one service execution parameter may be:

[0204] Calculate the difference between the execution state of the historical service and the preset target execution state corresponding to the historical service;

[0205] Based on the difference, adjust the initial value of the at least one service execution parameter to obtain the target value of the at least one service execution parameter.

[0206] Specifically, the target execution state corresponding to the historical service refers to the execution state that can be achieved when executing the historical service according to the execution strategy preset for the historical service.

[0207] In practical applications, in order to better determine the target service execution model and thus obtain the target value of the service execution parameter, the target execution state corresponding to the preset historical service can be obtained, then the execution state is compared with the target execution state, the difference between the execution state and the target execution state is calculated, and then based on this difference, the initial value of the at least one service execution parameter is adjusted to obtain the target value of the at least one service execution parameter.

[0208] Preferably, when the difference is greater than the difference threshold, the target value of at least one service execution parameter is set to the initial value of at least one service execution parameter minus a first preset value; when the difference is less than or equal to the difference threshold, the target value of at least one service execution parameter is set to the initial value of at least one service execution parameter plus a second preset value. For example, if the difference threshold is 5, the first preset value is 0.1, the second preset value is 0.2, and the initial value of the service execution parameter is 1: if the difference is 0.6, then the target value of the service execution parameter is 1 - 0.1 = 0.9; if the difference is 0.4, then the target value of the service execution parameter is 1 + 0.2 = 1.2.

[0209] In one or more embodiments of this embodiment, according to the execution status of the historical service, the specific implementation process of obtaining the target value of at least one service execution parameter by adjusting the initial value of at least one service execution parameter may also be:

[0210] According to the execution status of the historical service, measure the initial value of at least one service execution parameter according to a preset measurement rule to obtain a measurement value;

[0211] In the case where the measurement value is less than the measurement threshold, adjust the initial value of at least one service execution parameter, and execute the step of measuring the initial value of at least one service execution parameter according to a preset measurement rule according to the execution status of the historical service;

[0212] In the case where the measurement value is greater than or equal to the measurement threshold, determine the initial value of at least one service execution parameter as the target value of at least one service execution parameter.

[0213] Specifically, the preset measurement rule refers to a preset measurement method; the measurement value refers to the value obtained by measuring the initial value of the service execution parameter according to the preset measurement rule; the measurement threshold refers to the preset value used to measure the measurement value.

[0214] In practical applications, after obtaining the execution status of historical services, the initial values of at least one service execution parameter can be measured or evaluated based on the execution status of historical services according to a preset measurement rule, and the measurement values of the initial values of each service execution parameter are obtained. Further, the measurement values of the initial values of each service execution parameter are respectively compared with a measurement threshold, and then the initial values of the service execution parameters whose measurement values are less than the preset threshold are adjusted. It should be noted that the adjusted initial values of the service execution parameters cannot be directly used as the target values of the service execution parameters, and the adjusted initial values of the service execution parameters need to be continuously measured, that is, the step of measuring the initial values of the at least one service execution parameter according to the execution status of the historical service according to a preset measurement rule is executed; the initial values of the service execution parameters whose measurement values are greater than or equal to the measurement threshold are determined as the target values of the service execution parameters.

[0215] For example, there are two service execution parameters PY1 and PY2: the initial value of PY1 is "11", and the initial value of PY2 is "22". According to the execution status of historical services, the initial values "11" and "22" are measured according to a preset measurement rule, and the measurement value of the initial value "11" is 0.5 and the measurement value of the initial value "22" is 0.8. When the measurement threshold is 0.6, since 0.5 is less than 0.6, the initial value "11" of PY1 needs to be adjusted. At this time, the initial value of PY1 is adjusted to "18". Again, according to the execution status of historical services, the initial value "18" of PY1 is measured according to a preset measurement rule, and the measurement value of the initial value "18" is 0.7. At this time, 0.7 is greater than 0.6, and "18" is used as the target value of the service execution parameter PY1; since 0.8 is greater than 0.6, "22" is directly used as the target value of the service execution parameter PY2.

[0216] After obtaining the execution status of historical services, the initial values of all service execution parameters can also be measured based on the execution status of historical services according to a preset measurement rule, and a measurement value corresponding to the whole of the initial values of all service execution parameters is obtained. Further, the obtained measurement value is compared with the measurement threshold. In the case where the measurement value is less than the preset threshold, the initial values of each service execution parameter are adjusted. It should be noted that the adjusted initial values of the service execution parameters cannot be directly used as the target values of the service execution parameters, and the adjusted initial values of the service execution parameters need to be continuously measured, that is, the step of measuring the initial values of the at least one service execution parameter according to the execution status of the historical service according to a preset measurement rule is executed; in the case where the measurement value is greater than or equal to the measurement threshold, the initial values of each service execution parameter are determined as the target values of the service execution parameters.

[0217] For example, there are two business execution parameters PY3 and PY4. According to the execution status of historical business, the initial values of PY3 and PY4 are measured according to a preset measurement rule, and a measurement value 1 is obtained. When the measurement threshold is 0.9, since 1 is greater than 0.9, the initial values of each business execution parameter are the target values; when the measurement threshold is 1.5, since 1 is less than 1.5, the initial values of PY3 and PY4 need to be adjusted. Again, according to the execution status of historical business, the initial values of PY3 and PY4 are measured according to the preset measurement rule, and the measurement value obtained is 2. At this time, 2 is greater than 1.5, and the adjusted initial values of PY3 and PY4 are used as the target values.

[0218] It should be noted that although the execution environment of the business is constantly changing, adjusting the initial values of at least one business execution parameter as long as the execution environment of the business changes will greatly increase the data processing volume and reduce the efficiency of business execution. In addition, since the execution environment of the business has changed within a short period of time, but the degree of change is small, therefore, at intervals of a certain period of time, the initial values of at least one business execution parameter can be adjusted once. That is, according to the execution status of the historical business, the specific implementation process of adjusting the initial values of the at least one business execution parameter to obtain the target values of the at least one business execution parameter can be:

[0219] Obtain a preset time value;

[0220] At intervals of the time value, according to the execution status of the historical business, adjust the initial values of the at least one business execution parameter to obtain the target values of the at least one business execution parameter.

[0221] Specifically, the time value is a numerical value representing time, such as 1 minute, 1 hour, 1 day, etc.

[0222] Exemplarily, if the obtained preset time value is 5 minutes, then at intervals of 5 minutes, the initial values of at least one business execution parameter will be adjusted according to the execution status of historical business, so as to obtain the target values of at least one business execution parameter.

[0223] Step 106: Use the target business execution model to determine the execution strategy for executing the target business.

[0224] On the basis of constructing a target business execution model with pre-invested data and business target data as constraint conditions, further, use this target business execution model to determine the execution strategy for executing the target business.

[0225] Specifically, an execution strategy refers to a solution for executing a target business determined based on pre-investment data and business target data, with the goal of achieving the best effect or the best cost performance. It can also be the target investment data or the best investment data for the target business.

[0226] In practical applications, after determining the target business execution model, the execution strategy for executing the target business can be directly determined using the target business execution model, and then the target business can be executed according to this execution strategy.

[0227] For example, when the business target data of the target business is reached, based on the target business execution model, the target investment data for executing the target business can be determined, where the target investment data is less than or equal to the pre-investment data of the target business.

[0228] A business execution method provided in this specification obtains an execution request for a target business, where the execution request carries pre-investment data and business target data for executing the target business; then constructs a target business execution model with the pre-investment data and the business target data as constraint conditions; and further determines an execution strategy for executing the target business using the target business execution model. By establishing a target business execution model with the pre-investment data and business target data carried in the execution request of the target business as constraint conditions, the business is uniformly planned, and the efficiency of determining the execution strategy is improved to a certain extent, enabling a unified and optimal execution strategy during business execution. In addition, this application constructs a linear programming model and dynamically adjusts the initial values of business execution parameters according to the execution status of historical businesses, making the target business execution model as close to the optimal as possible and minimizing the impact of environmental fluctuations on the execution effect, thereby achieving the determination of the optimal execution strategy for different businesses in the same method.

[0229] The following combines the attached Figure 2 , taking the application of the business execution method provided in this specification in an online advertising scenario as an example, to further illustrate the business execution method. Among them, Figure 2 shows a process flow chart of a business execution method provided in an embodiment of this specification, which specifically includes the following steps.

[0230] Step 202: Obtain an execution request for the target advertising business, where the execution request carries pre-investment data and business target data for executing the target business.

[0231] Step 204: Construct a pre-investment execution algorithm that conforms to linear programming with the pre-investment data as the first constraint condition.

[0232] Step 206: Construct a business target execution algorithm that conforms to linear programming with the business target data as the second constraint condition.

[0233] It should be noted that Step 204 and Step 206 can be executed simultaneously; or Step 204 can be executed first and then Step 206; or Step 206 can be executed first and then Step 204. In this embodiment, the case where Step 204 and Step 206 are executed simultaneously is taken as an example for illustration.

[0234] Step 208: Construct a linear programming model according to the pre-investment execution algorithm and the business objective execution algorithm.

[0235] Among them, the linear programming model represents the constraint relationship between the business objective data and the pre-investment data.

[0236] Step 210: Identify the data type of the business objective data.

[0237] Step 212: Divide the business objective data into first-class data and second-class data according to the data type of the business objective data.

[0238] Among them, the first-class data is related to the pre-investment data, and the second-class data is not related to the pre-investment data.

[0239] Step 214: Construct an initial business execution model by using the linear programming model according to the first-class data and the second-class data.

[0240] Step 216: Obtain a preset time value.

[0241] Step 218: Obtain the execution status of the historical business at each interval of the time value.

[0242] Step 220: Calculate the difference between the execution status of the historical business and the target execution status corresponding to the preset historical business.

[0243] Step 222: Based on the difference, adjust the initial values of at least one business execution parameter to obtain the target values of at least one business execution parameter.

[0244] Step 224: At each interval of the time value, measure the initial values of at least one business execution parameter according to the execution status of the historical business according to the preset measurement rule to obtain a measurement value.

[0245] Step 226: In the case where the measurement value is less than the measurement threshold, adjust the initial values of at least one business execution parameter and execute Step 224.

[0246] Step 228: In the case where the measurement value is greater than or equal to the measurement threshold, determine the initial values of at least one business execution parameter as the target values of at least one business execution parameter.

[0247] Step 230: For each interval time value, based on the historical service execution status and the initial value of at least one service execution parameter, an evaluation result corresponding to the initial value of at least one service execution parameter is determined using an evaluation function.

[0248] Step 232: Determine an adjustment strategy for the initial value of at least one service execution parameter using a strategy function based on the evaluation result;

[0249] Step 234: According to the adjustment strategy, the initial value of at least one service execution parameter is adjusted to obtain a target value of at least one service execution parameter.

[0250] It should be noted that steps 218 to 222, 224 to 228, and 230 to 234 are three parallel schemes, all of which adjust the initial value of at least one business execution parameter based on the execution status of historical business, and then determine the target value of at least one business execution parameter. Different schemes can be selected according to needs.

[0251] Step 236: Update the initial service execution model according to the target value of at least one service execution parameter to obtain a target service execution model.

[0252] Step 228: Determine an execution strategy for executing the target business using the target business execution model.

[0253] This specification provides a business execution method, which obtains an execution request for a target business, wherein the execution request carries pre-investment data and business target data for executing the target business; then uses the pre-investment data and the business target data as constraints to construct a target business execution model; and then uses the target business execution model to determine the execution strategy for executing the target business. By using the pre-investment data and business target data carried by the execution request of the target business as constraints, a target business execution model is established, which realizes unified planning of the business and improves the efficiency of determining the execution strategy to a certain extent, so that a unified optimal execution strategy can be used when executing the business. In addition, this application constructs a linear programming model and dynamically adjusts the initial values of the business execution parameters according to the execution status of historical businesses, so that its target business execution model may be close to the optimal one, and avoids the influence of environmental fluctuations on the execution effect as much as possible, thereby realizing the determination of the optimal execution strategy for different businesses in the same method.

[0254] Corresponding to the above method embodiment, this specification also provides a service execution device embodiment, Figure 3 FIG1 shows a schematic diagram of the structure of a service execution device provided by an embodiment of this specification. Figure 3 As shown, the device includes:

[0255] An acquisition module 302, configured to acquire an execution request for a target service, where the execution request carries pre-investment data and service target data for executing the target service;

[0256] A construction module 304, configured to construct a target service execution model with the pre-investment data and the service target data as constraint conditions;

[0257] A determination module 306, configured to determine an execution strategy for executing the target service by using the target service execution model.

[0258] In one or more embodiments of this embodiment, the construction module 304 is further configured to:

[0259] Construct a linear programming model with the pre-investment execution data and the service target data as constraint conditions, where the linear programming model represents the constraint relationship between the service target data and the pre-investment data;

[0260] Construct an initial service execution model by using the linear programming model;

[0261] Adjust at least one service execution parameter of the initial service execution model according to a preset adjustment rule to obtain the target service execution model.

[0262] In one or more embodiments of this embodiment, the construction module 304 is further configured to:

[0263] Identify the data type of the service target data;

[0264] Divide the service target data into first-class data and second-class data according to the data type of the service target data, where the first-class data is related to the pre-investment data and the second-class data is not related to the pre-investment data;

[0265] Construct the initial service execution model by using the linear programming model according to the first-class data and the second-class data.

[0266] In one or more embodiments of this embodiment, the construction module 304 is further configured to:

[0267] Obtain the execution status of a historical service;

[0268] Adjust the initial value of the at least one service execution parameter according to the execution status of the historical service to obtain the target value of the at least one service execution parameter;

[0269] Update the initial service execution model according to the target value of the at least one service execution parameter to obtain the target service execution model.

[0270] In one or more embodiments of this embodiment, the building block 304 is further configured to:

[0271] According to the execution status of the historical service and the initial values of the at least one service execution parameter, use an evaluation function to determine an evaluation result corresponding to the initial values of the at least one service execution parameter;

[0272] According to the evaluation result, use a policy function to determine an adjustment policy for the initial values of the at least one service execution parameter;

[0273] According to the adjustment policy, adjust the initial values of the at least one service execution parameter to obtain the target values of the at least one service execution parameter.

[0274] In one or more embodiments of this embodiment, the building block 304 is further configured to:

[0275] Calculate the difference between the execution status of the historical service and the target execution status corresponding to the preset historical service;

[0276] Based on the difference, adjust the initial values of the at least one service execution parameter to obtain the target values of the at least one service execution parameter.

[0277] In one or more embodiments of this embodiment, the building block 304 is further configured to:

[0278] According to the execution status of the historical service, measure the initial values of the at least one service execution parameter according to a preset measurement rule to obtain a measurement value;

[0279] In the case where the measurement value is less than the measurement threshold, adjust the initial values of the at least one service execution parameter, and perform the step of measuring the initial values of the at least one service execution parameter according to the execution status of the historical service and the preset measurement rule;

[0280] In the case where the measurement value is greater than or equal to the measurement threshold, determine the initial values of the at least one service execution parameter as the target values of the at least one service execution parameter.

[0281] In one or more embodiments of this embodiment, the building block 304 is further configured to:

[0282] Obtain a preset time value;

[0283] At each interval of the time value, adjust the initial values of the at least one service execution parameter according to the execution status of the historical service to obtain the target values of the at least one service execution parameter.

[0284] In one or more implementations of this embodiment, the building module 304 is further configured to:

[0285] Taking the pre-investment data as a first constraint condition, constructing a pre-investment execution algorithm that conforms to linear programming;

[0286] Using the business target data as a second constraint condition, constructing a business target execution algorithm that conforms to linear programming;

[0287] The linear programming model is constructed according to the pre-investment execution algorithm and the business goal execution algorithm.

[0288] This specification provides a business execution device, which obtains an execution request of a target business, wherein the execution request carries pre-investment data and business target data for executing the target business; then uses the pre-investment data and the business target data as constraints to construct a target business execution model; and then uses the target business execution model to determine the execution strategy for executing the target business. By using the pre-investment data and business target data carried by the execution request of the target business as constraints, a target business execution model is established, which realizes unified planning of the business and improves the efficiency of determining the execution strategy to a certain extent, so that a unified optimal execution strategy can be used when executing the business. In addition, this application makes the target business execution model as close to the optimal as possible by constructing a linear programming model and dynamically adjusting the initial values of the business execution parameters according to the execution status of historical businesses, thereby avoiding the influence of environmental fluctuations on the execution effect as much as possible, and realizing the determination of the optimal execution strategy for different businesses in the same method.

[0289] The above is a schematic diagram of a service execution device according to this embodiment. It should be noted that the technical solution of the service execution device and the technical solution of the service execution method described above are based on the same concept. For details not described in detail in the technical solution of the service execution device, please refer to the description of the technical solution of the service execution method described above.

[0290] Figure 4 The block diagram of a computing device 400 according to one embodiment of the present disclosure is shown. Components of the computing device 400 include, but are not limited to, a memory 410 and a processor 420. The processor 420 is connected to the memory 410 via a bus 430, and a database 450 is used to store data.

[0291] The computing device 400 also includes an access device 440, which enables the computing device 400 to communicate via one or more networks 460. Examples of such networks include the Public Switched Telephone Network (PSTN), Local Area Network (LAN), Wide Area Network (WAN), Personal Area Network (PAN), or a combination of communication networks such as the Internet. The access device 440 may include one or more of any type of wired or wireless network interfaces (e.g., Network Interface Controller (NIC)), such as an IEEE 802.11 Wireless Local Area Network (WLAN) wireless interface, Worldwide Interoperability for Microwave Access (Wi-MAX) interface, Ethernet interface, Universal Serial Bus (USB) interface, cellular network interface, Bluetooth interface, Near Field Communication (NFC) interface, and so on.

[0292] In one embodiment of the present specification, the above components of the computing device 400 and Figure 4 other components not shown may also be connected to each other, for example, via a bus. It should be understood that Figure 4 the block diagram of the computing device shown is for illustrative purposes only and is not a limitation on the scope of the present specification. Those skilled in the art can add or replace other components as needed.

[0293] The computing device 400 can be any type of stationary or mobile computing device, including mobile computers or mobile computing devices (e.g., tablet computers, personal digital assistants, laptop computers, notebook computers, netbooks, etc.), mobile phones (e.g., smartphones), wearable computing devices (e.g., smartwatches, smart glasses, etc.) or other types of mobile devices, or stationary computing devices such as desktop computers or PCs. The computing device 400 can also be a mobile or stationary server.

[0294] Among them, the processor 420 is used to execute the following computer-executable instructions, which implement the steps of the above service execution method when executed by the processor.

[0295] The above is a schematic solution of a computing device according to this embodiment. It should be noted that the technical solution of this computing device and the technical solution of the above business execution method belong to the same concept. For the details not described in detail in the technical solution of the computing device, reference can be made to the description of the technical solution of the above business execution method.

[0296] An embodiment of this specification also provides a computer-readable storage medium, which stores computer-executable instructions. When the computer-executable instructions are executed by a processor, the steps of the above business execution method are implemented.

[0297] The above is a schematic solution of a computer-readable storage medium according to this embodiment. It should be noted that the technical solution of this storage medium and the technical solution of the above business execution method belong to the same concept. For the details not described in detail in the technical solution of the storage medium, reference can be made to the description of the technical solution of the above business execution method.

[0298] An embodiment of this specification also provides a computer program. When the computer program is executed on a computer, the computer is made to execute the steps of the above business execution method.

[0299] The above is a schematic solution of a computer program according to this embodiment. It should be noted that the technical solution of this computer program and the technical solution of the above business execution method belong to the same concept. For the details not described in detail in the technical solution of the computer program, reference can be made to the description of the technical solution of the above business execution method.

[0300] The above describes specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in a different order than in the embodiments and still achieve the desired result. Additionally, the processes depicted in the figures do not necessarily require the particular order or sequential order shown to achieve the desired result. In certain implementations, multitasking and parallel processing are also possible or may be advantageous.

[0301] The computer instructions include computer program code, which may be in the form of source code, object code, executable files, or some intermediate forms, etc. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. It should be noted that the content included in the computer-readable medium may be appropriately increased or decreased according to the requirements of legislation and patent practice within the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, the computer-readable medium does not include electrical carrier signals and telecommunication signals.

[0302] It should be noted that for the foregoing method embodiments, for the sake of simplicity of description, they are all expressed as a series of action combinations. However, those skilled in the art should know that the embodiments of this specification are not limited by the described action sequence, because according to the embodiments of this specification, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily essential to the embodiments of this specification.

[0303] In the above embodiments, the descriptions of each embodiment have their own focuses. For the parts not detailed in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0304] The preferred embodiments of this specification disclosed above are only used to help explain this specification. The optional embodiments do not elaborate on all the details and do not limit the invention to only the specific embodiments described. Obviously, many modifications and changes can be made according to the content of the embodiments of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the embodiments of this specification, so that those skilled in the art can well understand and utilize this specification. This specification is only limited by the claims and their full scope and equivalents.

Claims

1. A service execution method, comprising: Obtaining an execution request for a target service, the execution request carrying pre-investment data and service target data for executing the target service; Constructing a linear programming model with the pre-investment data and the business target data as constraints, wherein the linear programming model represents the constraint relationship between the business target data and the pre-investment data; Determine linear programming parameters using the linear programming model, and construct an initial business execution model based on the business execution parameters and the linear programming parameters; Adjusting at least one service execution parameter of the initial service execution model according to a preset adjustment rule to obtain the target service execution model, wherein adjusting at least one service execution parameter of the initial service execution model according to the preset adjustment rule to obtain the target service execution model includes: obtaining an execution status of a historical service; adjusting an initial value of the at least one service execution parameter according to the execution status of the historical service to obtain a target value of the at least one service execution parameter; and updating the initial service execution model according to the target value of the at least one service execution parameter to obtain the target service execution model; An execution strategy for executing the target service is determined using the target service execution model.

2. The method according to claim 1, wherein determining the service execution parameters by using the linear programming model and constructing the initial service execution model according to the service execution parameters comprises: Identify the data type of the business target data; According to the data type of the business target data, the business target data is divided into a first category of data and a second category of data, the first category of data is related to the pre-investment data, and the second category of data is not related to the pre-investment data; The initial business execution model is constructed based on the first category of data and the second category of data using the linear programming model.

3. The method according to claim 1, wherein adjusting the initial value of the at least one service execution parameter according to the execution status of the historical service to obtain the target value of the at least one service execution parameter comprises: Determining an evaluation result corresponding to the initial value of the at least one service execution parameter using an evaluation function according to the execution status of the historical service and the initial value of the at least one service execution parameter; Determining, based on the evaluation result, a strategy for adjusting the initial value of the at least one service execution parameter using a strategy function; According to the adjustment strategy, the initial value of the at least one service execution parameter is adjusted to obtain a target value of the at least one service execution parameter.

4. The method according to claim 1, wherein adjusting the initial value of the at least one service execution parameter according to the execution status of the historical service to obtain the target value of the at least one service execution parameter comprises: Calculating the difference between the execution status of the historical business and a preset target execution status corresponding to the historical business; Based on the difference, the initial value of the at least one service execution parameter is adjusted to obtain a target value of the at least one service execution parameter.

5. The method according to claim 1, wherein adjusting the initial value of the at least one service execution parameter according to the execution status of the historical service to obtain the target value of the at least one service execution parameter comprises: According to the execution status of the historical service and in accordance with a preset measurement rule, measuring the initial value of the at least one service execution parameter to obtain a measurement value; When the metric value is less than the metric threshold, adjusting the initial value of the at least one service execution parameter, and performing the step of measuring the initial value of the at least one service execution parameter according to the execution status of the historical service and the preset metric rule; In a case where the metric value is greater than or equal to the metric threshold, the initial value of the at least one service execution parameter is determined as the target value of the at least one service execution parameter.

6. The method according to any one of claims 4-5, wherein adjusting the initial value of the at least one service execution parameter according to the execution status of the historical service to obtain the target value of the at least one service execution parameter comprises: Get the preset time value; At intervals of the time value, the initial value of the at least one service execution parameter is adjusted according to the execution status of the historical service to obtain a target value of the at least one service execution parameter.

7. The method according to claim 1, wherein the linear programming model is constructed using the pre-investment data and the business target data as constraints, comprising: Taking the pre-investment data as a first constraint condition, constructing a pre-investment execution algorithm that conforms to linear programming; Using the business target data as a second constraint condition, constructing a business target execution algorithm that conforms to linear programming; The linear programming model is constructed according to the pre-investment execution algorithm and the business goal execution algorithm.

8. A service execution device, comprising: an acquisition module configured to acquire an execution request of a target service, wherein the execution request carries pre-investment data and service target data for executing the target service; a construction module configured to construct a linear programming model using the pre-investment data and the business target data as constraints, wherein the linear programming model represents a constraint relationship between the business target data and the pre-investment data; Utilizing the linear programming model, determining linear programming parameters, constructing an initial business execution model based on the business execution parameters and the linear programming parameters; adjusting at least one business execution parameter of the initial business execution model according to a preset adjustment rule to obtain the target business execution model, wherein adjusting at least one business execution parameter of the initial business execution model according to the preset adjustment rule to obtain the target business execution model includes: obtaining an execution status of a historical business; adjusting an initial value of the at least one business execution parameter according to the execution status of the historical business to obtain a target value of the at least one business execution parameter; and updating the initial business execution model according to the target value of the at least one business execution parameter to obtain the target business execution model; The determination module is configured to determine an execution strategy for executing the target service by using the target service execution model.

9. A computing device comprising: memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the business execution method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing computer-executable instructions, wherein the computer-executable instructions, when executed by a processor, implement the steps of the service execution method according to any one of claims 1 to 7.

11. A computer program product, comprising a computer program / instruction, which implements the steps of the service execution method according to any one of claims 1 to 7 when executed by a processor.

Images